The metabolic effects of 52 weeks treatment with the aldose reductase inhibitor ponalrestat were examined in 32 diabetic patients (16 insulin treated) in a randomized, double-blind, placebo-controlled clinical trial. Twelve hour metabolic profiles were performed on two separate occasions in each patient (a) during a single-blind placebo run-in period and (b) after 52 weeks treatment with either ponalrestat 600 mg/day or matching placebo. No effects attributable to ponalrestat were evident in glucose, pyruvate, or alanine metabolism. A significant overall treatment effect was observed for lactate concentration (ponalrestat vs. placebo 12 h least square mean at 52 weeks: 1.35 vs. 1.65 mmol/l, p = 0.024). For glycerol (p = 0.018), non-esterified fatty acids (p = 0.003) and total ketone bodies (p = 0.045) there was evidence for a variation of treatment with time between the insulin treated and non-insulin treated patients, although no statistically significant overall treatment effects were observed for any metabolite. Fasting total ketone body concentration at 52 weeks was significantly elevated in the insulin-treated patients receiving ponalrestat (antilog LS mean: 0.12 vs. 0.01 mmol/l, p = 0.01). In conclusion, ponalrestat has no effect on glucose metabolism in diabetic patients. A potentially beneficial effect on lactate metabolism was accompanied by a minor ketogenic effect in insulin-treated patients.