A population-based case-control study of lung cancer was performed in New Mexico to explain the differing patterns of lung cancer occurrence in the state's "Hispanic" Whites and other Whites. From 1980 through 1982, interviews were completed with 521 cases and 769 controls. In the male controls, the prevalence of current and previous cigarette usage was similar in the two ethnic groups, but Hispanics smoked fewer cigarettes daily. In the female controls, a lower percentage of Hispanics had ever smoked and their usual consumption was less than that of other White women. Older Hispanic female smokers had used hand-rolled cigarettes for an average of 8.8 years, whereas other White women of the same age had used this type for less than one-half year. Both stratified and multiple logistic analysis showed comparable risks of lung cancer in Hispanic White and other White smokers. There was no evidence of interaction between ethnicity and cigarette smoking. These analyses imply that the differences in lung cancer incidence between New Mexico's Hispanic Whites and other Whites are largely explained by the patterns of cigarette smoking of these two groups.