Posted by Ancestry Team on January 19, 2010 in Website

Last week we posted a new 1950 U.S. census substitute.  These records can serve as a great starting point for learning more about people who were born or grew up after the 1930 U.S. Census.

To protect the privacy of living individuals, U.S. censuses are not released to the general public until 72 years after the census was originally taken. Since the 1940 U.S. Federal Census will not be available until 2012 and the 1950 US Federal will not be made publicly available in 2022, city directories contain information about all the adult residents in a town, such as their name, occupation and home and work addresses.

The 1950 census Substitute searches across more than 2,500 U.S. city directories from the mid 1940s through the 1950s. City directories were precursors to modern-day phonebooks and contain the names of each adult resident in the town along with their occupation and home and work addresses. Until the 1950 U.S. Federal Census becomes available, these records serve as a great resource for finding any adult family members who would have been alive during the 1950s.

 1950 U.S. census substitute

You can learn more about how to use city directories  in our Learning Center.


  1. Ellen

    I don’t understand why you call these record compilations “Census Substitutes” and not just what they actually are — city directories. It’s misleading. I don’t call any set of records containing birthdates a “Birth Record Substitute”. It is confusing to see “Exact Search Results – 1950 Census Substitute” on one page and then “Exact Search Results – U.S. City Directories” on the next page. That will cause confusion down the line in sources. It sounds like a marketing tactic to take advantage of people’s impatience for more census records.

  2. If I follow the link for 1950s Census Substitute, and then put in a name (e.g., John Smith) and a place (e.g., San Diego, CA) I still get matches from 1897-1901 and on, not just matches from the 1940s and 1950s. It searches the larger City Directories set of records.

    I thought the 1950s Census Substitute (misnamed – I agree with Ellen on this) was supposed to isolate the results to the 1940s/50s.

    If I click on 1950s Census Substitute intentionally I shouldn’t have to wade through hundreds or thousands of matches to get to the 1950s.

  3. My experience has been the same as Randy Seaver’s (above, #2), so I’ll repeat his words:

    “If I click on 1950s Census Substitute intentionally I shouldn’t have to wade through hundreds or thousands of matches to get to the 1950s.”

    I also took the time to browse through the City Directories for some cities of interest to me, and I cannot find *any* new or recently added directories for those area that represent the 1950s (or 1930s or 1940s, for that matter).

    Is there a list of the relevant, newly-added directories? I sure can’t find it.


  4. Laura Dansbury

    #1 The components of the Census Substitute are clearly described as city directories. The information that can be obtained from them is similar to and overlaps the information in a census.

    #2 Randy, I cannot reproduce the issue you reported. Will you please email me directly with the URL of your results page?

    #3 Reed, I’m sorry you haven’t found anything new in the list of directories. We added hundreds of new city directories. I understand it would be helpful to know what has been updated and I will share your comment with the content team.

  5. Michele

    I would also like to see a list of what directories were added. Many of us check these databases regularly and it would help to know what cities/states were added.

    For Randy and Reed. I also had the same problem you mentioned with the search but if you put in the name, city and year you are interested in and select the box for “exact” you should get a relavent result. I usually don’t like to use this option but it helped me here.

  6. Andy Hatchett

    Ellen Re: #1

    I have been assured that the phrase “Census Substitute”, as used for the latest 1950s stuff, will be the last time it is used by Ancestry to describe such a collection.

    Ancestry management has come to the realization it is a misnomer.

  7. Andy Hatchett

    Laura Re: #4

    I would beg to differ with your point number one.

    Except for a name and address (and in some cases the name of a wife) there is *very* little genealogical information contained in these so-called substitutes.

    Luckily, we’ll see that phrase “Census Substitute” no more.

  8. Tom

    I did not find searching the “Census Substitute” very useful. It is the same as the other directories. One puts in a last name and receives hits regardless if the name of a street a business or whatever. I would have thought you could search by surname and find surnames.

  9. Michele

    I think the genealogical information varies depending on th directory you are looking at. I know the Greenville SC Directory lists the wife and the first initial of her maiden name, number of children and if the family owns the home or is renting or rooming there not to mention the occupation of the head of household and usually where he/she’s employed. The back of the directory usually has the names of people on each street and the house number so you can see their neighbors like you would on a census. While the layout is more difficult to navigate it can serve as a “census substitute” depending on the information you are looking for. The Greenville directory even designates the race if they are non-white.

  10. Jade

    Laura, re: #4, you say

    “The information that can be obtained from them is similar to and overlaps the information in a census.”

    Really not. No age, place of birth, parents’ places of birth, lists of children, number of years married, whether widowed or divorced (usually), date of immigration / naturalization, native language, how many children borne (and how many living) . . . .

    City Directories (and the similar regional ones) typically give street address, business address if owner or officer, occasionally name of spouse, frequently occupation and sometimes employer’s name, sometimes whether owning or renting.

    They are, as noted by others above, simply **not** “Census substitutes,” especially for the 1930s and before, when most people did not live in urban areas and were not listed in any City Directories. And because of regional development patterns, they *predominantly* cover the urban NE quarter of the USA, plus California, whereas the Census, especially post-1840, intends to cover the entire population in every location.

  11. Andy Hatchett

    Jade Re: #10

    I firmly believe that this term came from the Marketing People.

    Census Substitute sounds (to them) “sexier” than City Directories even tho it isn’t factual.

    Someday they’ll understand we want facts and help- not fluff and hype and that yes, we do know the difference.

  12. Susan

    I have the exact same problem as Randy Seaver. I type in my grandfather’s name and information, and I still get extraneous information that is completely outside the parameters I set in the information. Very frustrating.

  13. Laura Dansbury

    #7 & #10 have you found anything in city directories which helped you take the next step?
    #9, Michele, thank you for sharing your successes.

  14. Jason Thompson

    I also find the “Census Subsitute” to be absolutely infuriating to use.

    It seems that the majority (if not all) of these directories were not actually transcribed, but rather OCRed. My search results tend to include extraneous information, highly ranked, because two names happen to have been recognized anywhere in the same image.

    For example, if I search for John Smith, I might get a highly ranked result on an image that has “Seaver John” on the far left of the image and “Smith Francine” on the far right hand of the page.

    As punishment for your addition of this “Census Substitute”, I feel you employees should be banned from using Google for the next year and look up all your own searches in encyclopedias, phone books, yellow pages, card catalogs, and microfiche.

  15. Jade

    Laura, re: #13

    You ask regarding my #10, “have you found anything in city directories which helped you take the next step?”

    Heheheheh, since my ancestral researches are based on land, estate, court, tax, church, County vital and other records, there is no possibility that a city directory could have been of help (aside from the nonexistence of such for my ancestors living in rural PA, WV, NY and 17th/18th century PA, DE, WV, NY, CT, MA, ME, etc.).

    Oh there are some 19th-century directory entries found for very distant latter-day city-dwelling cousins. They are of slight interest for life-path details rather than genealogically helpful as to relationships. I found exactly two 20th-century entries for any ancestor (my father and his mother), who were not a genealogical problem for which the directory entries helped as ‘census substitute’. Since of course no relationship information was given in either entry (father boarding in town while at school; his mother living in another town with persons not identified as relatives), someone who did not already know who they were would not have found the entries genealogically useful.

    Your question was not about my point, which was that such directories are not “Census substitutes.” Whether they may provide interesting snippets is not at issue.

  16. Valerie

    For those having trouble with search results: I’ve found that I am not having much luck with the OCR search. For example, I was looking for Effie Craft in Birmingham, AL. If I do a search, I don’t receive any relevant results. However if I “flip” though the 1949 Birmingham Directory, I find her on page 206 (image 105). This is an issue with OCR and the technology behind it. If you know the city your ancestor lived in, I recommend “flipping” through the book.

    I too do not believe that this collection should be called a census substitute. First off, I love that there are new directories available and I understand that they will not supply the same information as a census. That doesn’t mean that they are not valuable to researchers. I’m more likely to learn an ancestor’s street address and employer through a directory than a census. I’m thankful for any new documents added.

    However… My issue is with the coverage of these directories. The census covers the entire country – these directories cover 2,500 cities. Hardly the same coverage as a census and thus not a true substitute.

  17. Shirley Nieminen

    I feel the reader comments have made a mess out of Laura Dansbury report. This should not be the place to vent any comments of personnel feelings. Just the short simple facts on trouble you had when you tried to use “Census Substitutes” or any name. All of you know what it ment or you would not have tried to use it. I wanted to get help but all of the complaints were so petty. Laura why don’t you do some editing and list only what went bad and if you know of a work around list just that in your answers. I know any new program will have problems and you do not have to take any heat. I would hate to have these coments make you to want to quite, hang in there Laura and edit, edit, edit…

  18. Andy Hatchett

    Shirley Re #19

    Having recently met Laura I can tell you she isn’t about to let a few negative comments (even mine *grin*) drive her away from a job that she loves; and that she loves it was very obvious in the two days I had interacting with her.

  19. Jerry Bryan

    I think the term “census substitute” is unfortunate, but it’s not just The entire genealogy community seems very cavalier about calling things census substitutes that really aren’t. Let’s give a break on this one.

    I would rather focus on making the searching better. For example, it’s not really possible to go into the Card Catalog and search for city directories for Knoxville, Tennessee (my home town). Rather, you have to find the city directories collection in the Card Catalog, and then specify Knoxville, Tennessee and the year as a part of your search.

    Then, New Search bites you. If you search for a surname in 1943 (for example), New Search gives you matches for other years as well. Old Search only gives you matches for 1943.

    And worst of all, the 1943 images for Knoxville, Tennessee only include the A’s and the B’s. The images for the C’s through the Z’s are missing completely. Lots of the city directories have this kind of missing image problem, and the problem has been reported before.

  20. Tara

    I’m a little disappointed. I thought the census was released every 70 years and was looking forward to using the 1940 census but I just read that they are now released every 72 years. Disappointed that I won’t be able to locate my mom on the census yet.

  21. Phyllis Sain

    Several times I have received an e-mail notice about the 1950 census substitute and when I open it there is nothing there.

  22. Jodi

    I am on the 14 day trial period, it has been a fun ride. But I have hit a brick wall, I need the 1950 census….NOT a “substitute”

    At this point…I have gone as far as I can go…on

  23. Laura Dansbury

    #14 & #17 – Yes, the city directories are OCR and not transcribed. OCR has challenges so this can make searching them different from searching a transcribed data collection. But the records do come online quicker.

    #17 – The tips for browsing are great. Even when we have transcribed records, I recommend spending time browsing the collection and understanding the context and looking for clues about relatives and neighbors before and after the record you are interested in.

    #19 & #20 – Thank you for the kind words and encouragement. I appreciate it.

    #21 – We are investigating the issue you reported. Thank you for the note.

    All – Thanks for the feedback on the naming of this collection. I can see there are strong opinions on this!

    I have found quite a variety of genealogic information including a count of the number of children under 16 in the household in the 1943 Knoxville directory. But it depends on the individual city directory.

  24. Andy Hatchett

    Jodi Re :# 24

    If you are looking for 1950 census stuff you’ve gone as far as you can go-period.

    Neither Ancestry nor any other organization will have that information until it is released to the public in 2022 as there is a law which forbids the release until 72 years after the census was taken.

  25. Jerry Bryan

    Re: #21 and the response in #25, let me provide some additional information.

    In New Search in exact mode, go to the U. S. City Directories collection and do a search on surname Bryan in 1943+-0. There are 33 matches – one for 1938, two for 1939

  26. Jerry Bryan

    Re: #21 and the response in #25, let me provide some additional information.

    In New Search in exact mode, go to the U. S. City Directories collection and do a search on surname Bryan in 1943+-0. There are 33 matches – one for 1938, two for 1939, one for 1942, and twenty-nine for 1943.

    It might seem counter-intuitive that there are twenty-nine matches for 1943, but the matches seem to be correct. That is, you might think there should be only one match. City directories are typically sorted by last name and first name, and any particular last name is typically listed only once followed immediately by all the first names. So you typically only expect one match per surname (unless the surname also appears as a first name, for example if there were a man named Bryan Smith, but I digress).

    The 1943 city directory for Knoxville also includes a directory of phone numbers sorted by phone number. The phone numbers for the Bryan surname therefore appear on a lot of different pages. Similarly, the 1943 city directory for Knoxville also includes a directory of addresses sorted by address. The addresses for the Bryan surname therefore appear on a lot of different pages. So far, so good.

    Now let’s look at the match for 1942. The match is on page 81, which is a telephone directory page embedded within the overall city directory, just like I was describing for 1943. There is a match for an R. O. Bryan in the rightmost column. That’s ok, I guess, except that this is a 1942 city directory and I specified 1943. But the more interesting fact is that there are bunch of matches in the middle column that have nothing to do with the Bryan surname.

    In particular, there is match for the character string “1933”, for the character string “1934”, etc. through the character string “1953”. These strings are not years, of course. They are parts of phone numbers. In general, matches such as these are an unavoidable and perfectly reasonable consequence of the fact that the images are being OCR’d and that the indexes are being generated automatically. should not be criticized for the occasionally oddity that happens like this as a part of the automated OCR indexing process.

    However, what does deserve criticism in this example is New Search. It appears to be looking for the string “1943” as a keyword in the content of the 1942 city directory rather than using “1943” to guide it to look only within the 1943 city directory. And not only that, it appears to be looking for 1943+-10 rather than 1943+-0 as I specified because it has matched on the character strings “1933” through “1953”.

    Old Search simply does not have this problem. It only gives you the twenty-nine matches for 1943.

    The New Search problem is a little hard to describe in general, but it afflicts New Search for many databases other than city directories. It’s like Old Search is a lot smarter than New Search (or New Search is a lot dumber than Old Search). That is, Old Search seems to understand when a year is a year and when an address is an address, etc., whereas New Search seems to be treating many search fields as keywords that match anywhere rather than treating the search fields as what they really are. As one example that is not city directories, New Search seems to treat addresses and phone numbers in the U.S. Public Records data base as keywords, whereas Old Search seems to treat addresses and phone numbers in the same database as addresses and phone numbers.

    A really good test for New Search is whether or not it can produce the same matches as does Old Search. In this case, New Search fails the test.

  27. Jerry Bryan

    Re: the second problem I raised in #21 and the response in #25, namely the missing pages. All the Knoxville city directories from the time had 1000+ pages. Here are the number of pages on

    1943 – 383 pages
    1942 – 1290 pages
    1941 – 17 pages
    1940 – 343 pages
    1939 – 338 pages
    1938 – 1222 pages
    1936 – 178 pages

    So of the seven years, only two of the city directories are of much value. For example, the 1943 directory includes surnames beginning with A and B, but not surnames beginning with C through Z. The few pages for 1941 are mostly just advertisements.

  28. I would like to second the detailed observations/comments of Jerry Bryan in #28 and #29.

    I have had similar problems searching City Directories, and I have also noticed missing pages in directories from Illinois, Wisconsin and other states.


  29. Peter Hudson

    It is interesting reading Jerry Bryan’s comments on New Searech v Old Search, and agree with all his points. As he says, Old Search seems very much smarter than New Search.

  30. Peter Hudson

    Bravo Jerry Bryan!

    It was interesting reading Jerry Bryan’s comments on New Search v Old Search. He has hit the nail on the head – Old Search is very much smarter than New Search! The most frustrating thing with New Search is how irrelevant the results are when carrying out a search on an individual. Such a search for an individual who’s date and place of birth is clearly stated will invariably produce results for potential ancestors that are wildly off the mark. It is also frustrating that forces New Search whenever a new log in is carried out.

    My biggest concern is that will take a decision that beauty (New Search) will finally be given preference over brains (Old Search), and put Old Search out to pasture. Ancestry appear to prefer to push a product that is aesthetically pleasing to them (despite their customers views) over a tried and properly tested one that has searved it very well indeed.

  31. Jerry Bryan

    Re: #32: It is also frustrating that forces New Search whenever a new log in is carried out.

    I don’t find this to be true. If I go to New Search and logoff, the system is in New Search when I log back on. If I go to Old Search and logoff, the system is in Old Search when I log back on.

    I try to use New Search as much as possible, and I use New Search about 2/3 of the time anymore. New Search still has a ways to go, but it has improved tremendously since it was first released.

    There is one option that New Search used to remember that it now seems to forget if I logoff and back on. This one drives me crazy, because I almost never want to see results displayed with the “Sorted by Relevance”. I almost always want to see them displayed with the “Summarized by category” option. If I set New Search to use “Summarized by category” and logoff, it seems to revert to “Sorted by Relevance” when I log back on.

    On the flip side, I always want to use New Search with “Advanced” enabled and with “Match all terms exactly” enabled. You can’t use “Match all terms exactly” unless “Advanced” is enabled. And New Search does remember these options for me across a logoff and log back on sequence. This is a really good thing.

  32. Laura Dansbury

    #29 & #30 The way these directories were originally filmed, a directory for a particular year may span two microfilm rolls – the first part of the directory appears at the end of one roll and the end of the directory appears at the beginning of the next roll. Unfortunately, this problem has been replicated in our browse tables. This means that you may need to look at two “Year” browse entries to view a complete directory. For example, the first part of a 1942 directory may appear in the browse entry labeled “1941-1942”. The second part of the directory would then appear in the browse entry labeled “1942-1943”.

    Due to this split, we have also found that sometimes we are missing part of a directory as we don’t have all microfilm rolls yet. We are working on producing these missing rolls and getting them added to the database to improve the browsing experience.

    In addition, we continue to work on our huge city directory collection and as it is completed it will naturally fill in these gaps. We are doing this in addition to the explicit effort to make the posted city directories more complete.

  33. Jade

    Laura, re: #25, you said:

    “All – Thanks for the feedback on the naming of this collection. I can see there are strong opinions on this!”

    This is not a “naming” problem, it is a misleading-marketing problem.

    I just received a “Monthly Update” in my message box that has plastered all over it, “1950 U.S. Census Substitute” . . . . “1950 U.S. Census Substitute” . . . . . “1950 U.S. Census Substitute” . . . .

    You know very well that this group of directories is nowhere close to being a “1950 U.S. Census Substitute,” not even close for the small minority of US-dwellers who are listed in the directories.

    This is a mean deception for newbies who haven’t worked back past the year 1900 and who have little idea of what the US Federal Census enumerations are really about.

    Why is it so hard to tell the truth?

  34. Andy Hatchett

    Jade Re: #35

    Right on!

    I will say this however. Things that are already in the pipeline are hard to stop- and this was already in the pipeline.

    I’ll be content if they just never use the phrase again for any future collection that really isn’t a census substitute.

    Hopefully later they will rename these two collections.

  35. Karl Bratcher

    I too, find the whole concept of City Directories in any configuration is unacceptable to me. Calling the 1940 & 1950 Census Substitutes an meaningful tool is completely out of order.

    I’ve been using FTM since its beginning and have had to put up with all the nickle and dime efforts along the way that I had to pay to this point. I still have a drawer full of all those CD research Disc that if I had the cost of them back I could buy a new car. This new idea I take as another Pacifier.

  36. walter ? pierce

    I am looking for 1/2 brother b in area around 1930 time .was raised by mother parents .might be a 1/2 sister adopted by name of Ann.
    Any records would be helpful

  37. Vicky

    I have had absolutely no luck with this set of records. When I put in a name to search, it uses that name on every record it returns regardless of whose name is really on the record. ie: I asked for Victor (last name)and every record I get for that last name says “Victor” until you actually go to the record; then the name might be Joe, Tom, John, etc. It won’t isolate to a specific State, much less City or Town!

  38. Sarah

    Question on 1930s and ’40s-era Polk Directory: Listed as the “County Taxpayer Directory,” each name has a dollar amount at the end of it. Is that taxable income or tax paid, or… ? (I can’t find any explanation for it in the directory itself.)

  39. Laura Dansbury

    #39 – If you are interested in browsing to a particular city, I suggest going to and using the right side bar navigation to narrow down to just the cities that you are interested in. We may not have the city you are interested in included in the 1950 era list, but we may have that city for a different year.



  41. Mary M.

    I found the city directories useful in helping establish the approximate death year if the person died after the 1930 census and before the time the SSDI kicks in; occasionally, no one in the family knows when the person in question died or where he or she is buried. Having a range of a couple years helps save me time when looking through newspaper microfilms for obits. No, directories are no substitute for the census but they do help establish residency, specific place of employment, the address & even some family members (by address). Generally, these directories also do a much, much better job with “foreign” names than do census takers & (especially) census transcribers.

  42. Laura Dansbury

    #42-The search logic is the same, regardless of subscription level. We veil some of the information such as city and date if you are not a subscriber.
    #43 – Thank you for sharing your tips & successes.

  43. Ray Phillips

    You U.S. types do have it cushy. In England, Scotland and Wales we have to wait 100 years NOT 72 as with you lot, for the results of a Census. At 69, I have just seen my parents (both born 1906) appear on the 1911 Census which was released (amidst much opposition) 2 years early.

  44. Doris Snyder

    Don’t care what you call them. If it helps with my research I appreciate it. “By any other name a rose is still a rose.”

  45. Doris Re: #46

    If a rose is a rose is a rose-
    Then why try and pass it off as a Lily?

    Things are called things for a reason- namely to have a common understanding of the thing so that communication is possible.

  46. Jerry Bryan

    Re: #41: The browsing method you suggest is generally speaking the best way to find things in city directories. However, I would suggest a slightly more intuitive way to get to the City Directory page itself.

    Namely, go to the Ancestry Card Catalog, for example, by hovering over the search button and then going down to click on the Ancestry Card Catalog. Once in the Ancestry Card Catalog, search on something like “city direct*” without the quotes. I tend to use the wild card in this context for fear of not knowing whether it’s spelled “Directory” or “Directories”. Then, click the first entry in the list.

    It’s really much simpler than it sounds, and for the most part is very intuitive. The hardest part if you are not used to it is the part about hovering over the search button because it is not at all obvious in advance that that’s the way to find the Ancestry Card Catalog. The rest of the process is very natural.

    (Dare I point out that searching for a title of “1950 census substitute” in the Ancestry Card Catalog will find nothing (grin!), although it does work as a keyword search.)

Comments are closed.