In preparation for islet transplantation, diabetes was induced using streptozotocin (STZ) in non-human primates ranging from juveniles to adults with diverse body types: we studied the process with respect to the diabetic state and emergence of adverse events (AEs) and their severity, and identified risk factors for clinical and laboratory AEs. Pharmaceutical-grade STZ was given based on body surface area (BSA) (1050-1250 mg/m(2), equivalent to 80-108 mg/kg) or on body weight (BW) (100 mg/kg) to 54 cynomolgus and 24 rhesus macaques. AEs were related to risk factors, i.e. obesity parameters, BW and BSA, age and STZ dose in mg/m(2). Clinical AEs during the first days after infusion prompted euthanasia of three animals. Except for those three animals, diabetes was successfully induced as shown by circulating C-peptide levels, the intravenous glucose tolerance test and/or arginine stimulation test. C-peptide after infusion weakly correlated (P = 0.048) with STZ dose in mg/m(2). Grade ≥3 nephrotoxicity or hepatotoxicity (serum markers >3× baseline or >5 × baseline, respectively) occurred in about 10% of cases and were generally mild and reversible. Grade ≥2 clinical AEs occurred in seven of 78 animals, reversed in four cases and significantly correlated with obesity parameters. Taking girth-to-height ratio (GHtR) as an indicator of obesity, with threshold value 0.92-0.95, the positive predictive value of obesity for AEs was 92% and the specificity 94%. We conclude that diabetes is successfully induced in non-obese animals using a 100 mg/kg pharmaceutical grade STZ dose. Obesity is a significant risk factor, and animals with a higher than normal GHtR should preferably receive a lower dose. The incidence of relevant clinical or laboratory AEs is low. Careful monitoring and supportive medical intervention can result in recovery of AEs.