To determine the incidence of type 2 diabetes in people with a history of hospitalization for major mental illness versus no mental illness in Scotland by time period and sociodemographics.We used national Scottish population-based records to create cohorts with a hospital record of schizophrenia, bipolar disorder, or depression or no mental illness and to ascertain diabetes incidence. We used quasi-Poisson regression models including age, sex, time period, and area-based deprivation to estimate incidence and relative risks (RRs) of diabetes by mental illness status. Estimates are illustrated for people aged 60 years and in the middle deprivation quintile in 2015.We identified 254,136 diabetes cases during 2001-2015. Diabetes incidence in 2015 was 1.5- to 2.5-fold higher in people with versus without a major mental disorder, with the gap having slightly increased over time. RRs of diabetes incidence were greater among women than men for schizophrenia (RR 2.40 [95% CI 2.01, 2.85] and 1.63 [1.38, 1.94]), respectively) and depression (RR 2.10 [1.86, 2.36] and 1.62 [1.43, 1.82]) but similar for bipolar disorder (RR 1.65 [1.35, 2.02] and 1.50 [1.22, 1.84]). Absolute and relative differences in diabetes incidence associated with mental illness increased with increasing deprivation.Disparities in diabetes incidence between people with and without major mental illness appear to be widening. Major mental illness has a greater effect on diabetes risk in women and people living in more deprived areas, which has implications for intervention strategies to reduce diabetes risk in this vulnerable population.