What changed across decades
The US federal census has been taken every ten years since 1790. Early schedules (1790–1840) name only the head of household; other members are tallied by age bracket. From 1850 onward, every free person in the household is named. 1880 introduced the relationship-to-head-of-household column, which is the single most valuable field for genealogy. 1890 is almost entirely lost to a 1921 fire; expect a gap. 1900 adds month and year of birth, marriage duration, and immigration year. 1920 and 1930 add mother-tongue and naturalization status fields that are gold for tracking immigrants.
State censuses — New York, Iowa, Kansas, Wisconsin — fill in the between-decade years and often carry fields that federal schedules lacked.
Abbreviations in the enumerator's hand
Relationship codes: Hd. (head), Wf. or Ux. (wife), Dau. (daughter), Son, Br. (brother), Sis. (sister), F-in-L (father-in-law). Race columns use W, B, Mu (mulatto), Chi, Jap, In (American Indian) depending on the era. Occupation columns lean on period-specific shorthand — see the occupations guide.
Marital status: S (single), M (married), Wd. (widowed), D (divorced). A checkmark or X in the 'can read' / 'can write' columns means yes, a blank means no — not the reverse.
Reading between the lines
Enumerators walked a defined route, so neighbors on the page are usually geographic neighbors in real life. Look at pages before and after your family — in-laws and siblings often lived next door. Mismatched ages across decades usually reflect the informant's guess, not a different person; the key anchor is the relationship column combined with birthplace.