Every number in this chart comes from the U.S. Census Bureau. This page sets out where each number comes from, how each category is defined, what is interpolated and what to be careful about.
It is a bar-chart race of New York City's population by race and ethnicity at each decennial census from 1900 to 2020. The five-borough City of New York was consolidated in 1898, so 1900 is the first census of the city in its modern boundaries — that is why the series starts there and why the geography stays essentially constant across all 13 censuses.
Two views are available:
Race-alone categories. White, Black, Asian and Pacific Islander, and Other or multiracial are reported as "race alone." They sum to about 100% in every year. "Other or multiracial" bundles American Indian and Alaska Native, "some other race," and (from 2000, when the census first allowed it) people reporting two or more races.
Non-Hispanic White. The data table below also lists non-Hispanic White (White and not Hispanic), which the census reports for 1940 and continuously from 1970. It is not drawn as a separate bar on the chart, because the difference between race-alone White and non-Hispanic White is only the small slice of Hispanic New Yorkers who mark their race as White — most mark “some other race” or two or more races instead, so that slice is far smaller than the full Hispanic population and showing it on the White bar was more confusing than clarifying. It is kept in the table for reference.
Counts. Person counts are derived from each group's published percentage multiplied by the census total for that year: count = (share / 100) × total. Because the published shares are rounded to two decimals, a derived count can differ from the Census Bureau's own published count by a small rounding amount (at most a few hundred people in the early, tiny categories). The city totals themselves are the published census totals.
Interpolation. A census is taken once a decade. To animate the chart, values between two censuses are linearly interpolated, so the bars glide rather than jump. The figures are exact only at each decade (when the year reads 1900, 1910, … 2020). Pause on a census year, or read the methodology table below, for the real numbers. A group that the census had not yet counted grows in from zero at the decade it first appears.
Between 2010 and 2020 the race-alone White share falls sharply (44.0% to 34.1%) and "Other or multiracial" jumps (17.7% to 28.1%). Part of this is real change, but part is a methodological shift: in 2020 the Census Bureau changed how it coded write-in race responses, and far more people — especially Hispanic respondents — were classified as "some other race" or "two or more races" rather than White. The non-Hispanic White share, which is less affected, fell much more gently (33.3% to 30.9%). Read the 2010-to-2020 race-alone change with that in mind.
Shares as used by the chart (race alone, except where noted). The Hispanic and non-Hispanic White columns overlap the race columns and do not add into the 100%.
| Year | Total | White | Black | Asian/PI | Other | Hispanic | NH White |
|---|---|---|---|---|---|---|---|
| 1900 | 3,437,202 | 98.04 | 1.76 | 0.19 | 0.00 | — | — |
| 1910 | 4,766,883 | 97.95 | 1.92 | 0.12 | 0.01 | — | — |
| 1920 | 5,620,048 | 97.14 | 2.71 | 0.14 | 0.00 | — | — |
| 1930 | 6,930,446 | 95.08 | 4.73 | 0.19 | 0.01 | — | — |
| 1940 | 7,454,995 | 93.59 | 6.15 | 0.24 | 0.01 | 1.62 | 91.97 |
| 1950 | 7,891,957 | 90.17 | 9.47 | 0.27 | 0.08 | — | — |
| 1960 | 7,781,984 | 85.33 | 13.98 | 0.55 | 0.13 | — | — |
| 1970 | 7,894,862 | 76.62 | 21.13 | 1.20 | 1.06 | 16.20 | 62.95 |
| 1980 | 7,071,639 | 60.72 | 25.23 | 3.27 | 10.77 | 19.88 | 51.88 |
| 1990 | 7,322,564 | 52.26 | 28.71 | 7.00 | 12.02 | 24.36 | 43.20 |
| 2000 | 8,008,278 | 44.66 | 26.59 | 9.90 | 18.85 | 26.98 | 34.98 |
| 2010 | 8,175,133 | 44.00 | 25.55 | 12.77 | 17.68 | 28.58 | 33.31 |
| 2020 | 8,804,190 | 34.09 | 22.08 | 15.73 | 28.11 | 28.29 | 30.89 |
Shares are percentages of the total city population. Em dashes mark years the census did not separately tabulate that group.