Analysis of a chronological trend in data from the second National Health and Nutrition Examination Survey indicated that average blood lead levels in the United States dropped approximately 37 per cent (5.4 micrograms per deciliter) from February 1976 through February 1980. There was no evidence that this trend was due to errors in laboratory measurement or to the design of the survey. The trend was present even after accounting for differences in race, sex, age, region of the country, season, income, and degree of urbanization. Changes in exposure to lead in paint or in the diet are unlikely explanations of the trend. However, the correlation of blood lead levels with the lead level in gasoline was highly significant (P less than 0.001) overall and in population subgroups defined by race, sex, and age. Although strong correlation does not prove cause and effect, the most likely explanation for the fall in blood lead levels is a reduction in the lead content of gasoline during this period.