Skip navigation

Card files are yesteryear’s version of databases and spreadsheets – and they were everywhere. Think of grandmother’s recipe card boxes, library card catalogs, and the executive’s “Rolodex” of important phone numbers. Federal agencies created hundreds of card files that are now in the National Archives as cards or on microfilm, including First Lady Bess Truman’s recipe file with Washington Cream Pie:

Image: Bess Wallace Truman’s “Recipe Washington Cream Pie” (NAID 139308685).


Early computers used “punch cards” for their data feed.  In the photo detail below, women “punch card operators” were compiling data from the original 1940 census schedules onto punch cards that would then be “run” through computer tabulation equipment to obtain compiled statistical data for subsequent published reports.

Image:  Detail from “Card Punch Operators working on agricultural cards. Population and housing cards carried 45 columns. All other cards carried 80 columns.” Local Identifier 29-C-1B-41 (NAID 6200858).  From series: “Photographs Documenting the Sixteenth Decennial Census, 1940-1941” (NAID 513293).


After the compilation of data from the 1950 census was completed, the Census Bureau’s Geography Division created a card file with basic 1950 census population information that would serve as a quick reference file.  During the 20th century, the Geography Division determined the boundaries of enumeration districts based on the political boundaries of counties, towns, townships, villages, and other “Minor Civil Divisions” (MCDs), as well as known changes in population since the previous census.  After the 1950 census was over, they would start planning for the 1960 census.


The Geography Division’s Card File of Population Data Relating to the 17th Census, 1950 (NAID 2990400) contains one card for each state except that the following are grouped together on a single card: (1) the New England States; (2) Maryland, Virginia, and the District of Columbia; (3) Kentucky and Tennessee; and (4) Minnesota, North Dakota, and South Dakota.  There are summary cards for the United States and for U.S. Territories. The cards answer these basic questions about each state’s population in 1950:


  • How many counties did the state have (or parishes in Louisiana)?
  • How many Minor Civil Divisions (MCDs) did the state have in the 1950 census?  MCDs were townships, towns, villages, and other units of local government.
  • What was the state’s total population in millions?
  • How many cities with over 250,000 population did the state have, and what was their population total in millions?
  • How many cities with between 50,000 and 250,000 population did the state have, and what was their population total in millions?


For example, the card for Ohio, shown below, indicates the state had 88 counties with 1,447 Minor Civil Divisions. Its 1950 population was 7.9 million. It had 6 cities with over 250,000 people each whose total population was 2.6 million. In addition, it had 8 cities with between 25,000 and 250,000 people each whose total population was .7 million (700,000).


Image:  Ohio Population, 1950 (NAID 195936142).


Today, U.S. population statistics for 1950 and other years can easily be found online, although the speed with which information is retrieved may depend on our search terms, preferred search engine, and other factors. This record series reminds us of the important part that card files played in records storage and retrieval before the computer era.

Locating family members in the 1790 to 1940 census records can sometimes be challenging. These 20 tips for research success may help avoid frustration.


  1. Start with yourself. Organize what you know on an ancestral chart as well as on family group sheets for each ancestral couple. Write in pencil! Your knowledge will change and grow during your research.
  2. Start with ancestors who were alive in 1940. Find them on the 1940 census, and then work backwards to 1930, 1920, 1910, and so forth.... Expand your knowledge base by searching for ancestor’s siblings and other relatives. The more information you have about a family group, the more successful you will be in locating them in previous censuses.
  3. Use the clues (names, ages, birthplaces, occupations, relationships to other persons in the household) in one census to locate the same individuals in earlier censuses.
  4. Names, ages, birthplaces, and other information may not be 100% correct on each census. Accuracy depended on the knowledge or memory of the person providing the information. Only the 1940 census identifies which household member provided the information with an “X in a circle.”
  5. Names may be spelled differently than you expect. The enumerator may have written the name down according to his or her idea of how the name was spelled. The enumerator may not have asked how the name was spelled, and the family member answering the questions may not have been able to spell their own name. Some people are listed by initials, such as T. A. Smith, or by abbreviations, such as “Thos.” for “Thomas.” Think creatively!
  6. Many people had similar names so it is necessary to sort them out by other personal details, such as age, birthplace, family composition, and so forth.
  7. Some people were listed under erroneous names by mistake. For example, children of remarried widows may be listed under the stepfather’s surname instead of their own surname. Recent immigrants’ names were sometimes garbled by enumerators who had difficulty understanding foreign accents. Blatantly fictional names for real people were rare but were occasionally reported.
  8. The birthplace of foreign-born persons may reflect the person’s understanding of their “country” at the time they immigrated or at the time the census was taken. National boundaries were very fluid. Common generic “catch-alls” include “Great Britain” (instead of England, Scotland, Ireland, or Wales) or “Germany” (instead of specific principalities).
  9. Don’t ignore nearby neighbors:  they may be relatives, close friends, or neighbors from your own family’s previous place of residence. Related and unrelated families from one geographic area sometimes migrated to a new location where people they knew already lived.
  10. Most of the 1890 census was destroyed by fire, so the 20-year leap from 1900 back to 1880 may be hard to “jump.” The more information you collect about the entire extended family in the 1900 and later censuses can help you jump back to 1880. In addition, the 1890 “Special Schedules of the Eleventh Census Enumerating Union Veterans and Widows of Union Veterans of the Civil War” indicates where these individuals were living and information about their military service.
  11. Census records provide valuable clues to locate a person in earlier censuses and other records, including an approximate date of birth (1850-1940), place of birth (1850-1940), place of each parent’s birth (1880, 1900-1930), date of marriage (1900), number of children born to each woman (1900-1910), year of immigration to the U.S. (1900-1930), military service (1910, 1930), and ownership of real estate (1850-1940). Read Clues in Census Records, 1790-1840 and Clues in Census Records, 1850-1930 for more information. Carefully read census information so you don’t miss any valuable clues.
  12. The mortality schedules of 1850, 1860, 1870, and 1880—lists of persons who died during the year before the taking of the census—may provide information about family members who would otherwise have been omitted from the census because they were deceased.
  13. Check the agricultural schedules of 1850, 1860, 1870, and 1880 for farmer ancestors since these give interesting statistical data on the kinds and numbers of livestock, and the types and amount of food products grown on the farm during the year preceding the taking of the census.
  14. Check the manufacturing schedules of 1850, 1860, 1870, and 1880 for ancestors who engaged in businesses such as tanning, milling, coopering, or cheese making, among many others. Farmers often had sideline manufacturing businesses, too, so don’t ignore these if your ancestor’s occupation was listed as “farmer” on the population schedule.
  15. Census takers wrote down a person’s race based on his or her visual assessment of the person, so that variations in a person’s “race” (white, black, mulatto) over time is not unusual.
  16. Know the official census day! The information collected was supposed to be accurate as of the official census day, not the date the enumerator visited the household. These dates were August 2, 1790; August 4, 1800; August 6, 1810; August 7, 1820; June 1, 1830; June 1, 1840; June 1, 1850; June 1, 1860; June 1, 1870; June 1, 1880; June 1, 1890; June 1, 1900; April 15, 1910; January 1, 1920; April 1, 1930 (Oct. 1, 1929, in Alaska); and April 1, 1940.
  17. If you have trouble finding one ancestor, don’t give up researching! Instead, search for a different related person that likely lived in the same household or nearby or in the same community. You may find the first person you were looking for by that “back door” approach.
  18. Document your sources. For census records, details that will help you return to the same census record--or share it with another person--include details such as the census year, town, county, state, enumeration district number, and page (sheet) and line numbers.
  19. Don’t give up if the search engine “can’t find” the name or person you’re looking for. Don’t rely solely on online search engines. The data may not have been transcribed accurately. If it’s accurate, it may not be what you expect. Be prepared to read census records page by page, line by line to look for variations in names and information that simple searches miss (you can do this online, too).
  20. The 1950 census will be released on April 1, 2022. Start getting ready by compiling lists of ancestors and their places of residence.