OBJECTIVES: Diabetes prevalence has increased markedly in Mexico. We examined the individual and joint contributions of economic disadvantage during childhood (EDDC) and elevated body weight on diabetes prevalence in 3 cohorts of Mexican adults. METHODS: Data on those 60-69 years old from the 1930-1939, 1940-1949, and 1950-1959 birth cohorts in Waves 1 (2001), 3 (2012), and 5 (2018) of the Mexican Health and Aging Study were used. EDDC was defined as the absence of a toilet in the household before age 10. Body mass status was defined using self-reported perceived body image at age 50. Diabetes was based on respondent reports. Supplementary analyses using HbA1c as a criterion for diabetes were conducted. A regression-decomposition approach was implemented. Logistic regression models included adjustments for sociodemographic characteristics and access to medical care. RESULTS: Diabetes prevalence was 23% overall and 11%, 25%, and 26% in the 1930-1939, 1940-1949, and 1950-1959 cohorts, respectively. EDDC declined across successive cohorts, whereas the prevalence of overweight/obesity at age 50 increased. EDDC and overweight/obesity were associated with higher odds of reporting diabetes. A scenario that eliminates disadvantaged EDDC reduced diabetes prevalence by 11% in a pooled sample, while eliminating overweight/obesity reduced it by 30%. Overweight/obesity explained 42% of the rise in diabetes prevalence between the 1930-1939 and 1950-1959 cohorts. Improvement in EDDC explained 18% of the rise in diabetes prevalence between 1930-1939 and 1950-1959 cohorts. DISCUSSION: High body weight across Mexican birth cohorts seemed to offset the potential benefits from improvements in childhood conditions on adult diabetes risk.