Factors were examined that might explain reported ethnic differences in leisure time physical activity (LTPA) between Mexican Americans (MAs) and European Americans (EAs). Data were from a random sample of 749 community-dwelling MAs and EAs, aged 65 and older, who participated in the San Antonio Longitudinal Study of Aging (SALSA) baseline examination. Variables examined included LTPA measured as kilocalories of energy expended per week, contextual variables (age, sex, socioeconomic status (SES), acculturation/structural assimilation), psychosocial measures (self-esteem, mastery, perceived health control), lifestyle variables (fat avoidance, current alcohol drinker, years smoking, body mass index (BMI)), and presence of chronic diseases (diabetes mellitus, angina pectoris, myocardial infarction, stroke, hypertension, arthritis, chronic obstructive pulmonary disease, depression, mild cognitive impairment). Hierarchical multiple regression was used to examine potential mediators of the ethnic group-LTPA association. EAs expended almost 300 kcal/wk more energy than did MAs (1,287 kcal/wk vs 1,001 kcal/wk). SES and psychosocial (self-esteem), lifestyle (fat avoidance, smoking, BMI), and disease (depression) factors that vary by SES explained this ethnic difference. In MAs, structural assimilation, but not acculturation, was significantly associated with LTPA independent of SES. Self-esteem, BMI, and depression explained this association. Psychosocial resources, lifestyle behaviors, and depression explain differences in LTPA between older MAs and EAs. Interventions to increase LTPA in both ethnic groups should be targeted especially at women and persons who have low self-esteem, smoke, and are obese or depressed. In MAs, additional emphasis should be focused on those who are less structurally assimilated into the broader American society.