Although obesity has been associated with an increased cancer risk in the general population, the association in patients with type 2 diabetes (T2D) remains controversial. We conducted a dose–response meta-analysis of cohort studies of body mass index (BMI) and the risk of total and site-specific cancers in patients with T2D. A systematic literature search was conducted in PubMed, Scopus, and Medline until September 2020 for cohort studies on the association between BMI and cancer risk in patients with T2D. Summary relative risks (RRs) and 95% confidence intervals (CIs) were calculated using random effects models. Ten prospective and three retrospective cohort studies (3,345,031 participants and 37,412 cases) were included in the meta-analysis. Each 5-unit increase in BMI (kg/m2) was associated with a 6% higher risk of total cancer (RR: 1.06, 95% CI 1.01, 1.10; I2 = 55.4%, n = 6), and with a 12% increased risk in the analysis of breast cancer (RR: 1.12, 95% CI 1.05, 1.20; I2 = 0%, n = 3). The pooled RRs showed no association with prostate cancer (RR: 1.02, 95% CI 0.92, 1.13; I2 = 64.6%, n = 4), pancreatic cancer (RR: 0.97, 95% CI 0.84, 1.11; I2 = 71%, n = 3), and colorectal cancer (RR: 1.05, 95% CI 0.98, 1.13; I2 = 65.9%, n = 2). There was no indication of nonlinearity for total cancer (Pnon-linearity = 0.99), however, there was evidence of a nonlinear association between BMI and breast cancer (Pnon-linearity = 0.004) with steeper increases in risk from a BMI around 35 and above respectively. Higher BMI was associated with a higher risk of total, and breast cancer but not with risk of other cancers, in patients with T2D, however, further studies are needed before firm conclusions can be drawn.