Aims/hypothesis: We assessed whether low birthweight or early adiposity rebound was more strongly associated with type 2 diabetes, and whether any effect of low birthweight or early adiposity rebound was explained by adult BMI, adult height, social class of subject or of his/her father, or maternal or paternal diabetes. Methods: Cox's proportional hazard models were used on data from the National Birth Cohort Study (the MRC National Survey of Health and Development), which was begun in 1946 and had self-reported physician-diagnosed diabetes with age at onset ranging from 31 to 53 years (n=78 cases, and n=47 cases in the multivariate analysis) as the outcome. Results: A U-shaped association between birthweight and type 2 diabetes rates was close to statistical significance (quadratic term p value=0.08). Younger age at adiposity rebound was associated with increased rates of type 2 diabetes (test for trend p=0.002), the association being robust to adjustment for each of sex, birthweight, weight at 2 years, father's social class, parental diabetes, and own social class. The effect of early adiposity rebound was very slightly reduced by adjustment for sex and adult height (p=0.003), but considerably reduced after adjustment for sex and adult BMI (test for trend p=0.1), and further reduced (p=0.4) after additional adjustment for birthweight, weight at 2 years, adult height, social class of subject and of his/her father, and parental diabetes. Conclusions/interpretation: Early adiposity rebound was associated with an increased rate of type 2 diabetes independently of birthweight, but its effect was mostly through high adult BMI. Parental diabetes and possibly low weight at 2 years were also risks.