We examined the associations between childhood and adult socioeconomic position (SEP) and incident diabetes in 7,432 individuals aged 50 or older from the English Longitudinal Study of Ageing (ELSA). We identified 174 and 189 cases of incident diabetes, in men and women, respectively, over 5.3 years of follow-up. Cox models were estimated. In women, childhood SEP, education, occupational class, income, wealth, and subjective social status (SSS) were related to incident diabetes. Occupational class, income, and SSS did not remain significantly related to incident diabetes after adjustment for individual sets of covariates (i.e. unhealthy behaviours, obesity, or psychosocial factors). Wealth (HR: 1.65, 95 % CI: 1.05, 2.60, poorest vs. wealthiest tertile) remained significantly related to incident diabetes after adjustment for all covariates, but education (HR: 1.46, 95 % CI: 0.92, 2.33, lowest vs. highest category) and childhood SEP (HR: 1.47, 95 % CI: 0.98, 2.19, lowest vs. highest category) did not. In men, only wealth and SSS were related to incident diabetes. SSS remained significantly related to incident diabetes after adjustment for all covariates (HR: 2.46, 95 % CI: 1.32, 4.68, lowest vs. highest category), but wealth did not (HR: 1.42, 95 % CI: 0.94, 2.15, poorest vs. wealthiest tertile). Additional adjustment for wealth did not greatly affect the association between incident diabetes and SSS in men. Incident diabetes in older women is associated with SEP from all life stages, while in older men only with current SEP. Psychosocial factors (in women), unhealthy behaviours, and obesity partly mediate these associations.