This study investigated the effect of varying degrees of renal insufficiency on the pharmacokinetics of rosiglitazone. Subjects were stratified by estimated creatinine clearance: normal (> 80 mL/min; n = 12), mild renal insufficiency (60-80 mL/min; n = 15), moderate renal insufficiency (30-59 mL/min; n = 18), and severe renal insufficiency not requiring dialysis (< or = 29 mL/min; n = 12). Plasma rosiglitazone concentrations and protein binding were determined after a single oral 8-mg dose of rosiglitazone. Total and unbound pharmacokinetic parameters were generated using noncompartmental methods. AUC, Cmax, and t1/2 data were analyzed separately by ANOVA to provide point estimates and corresponding 95% confidence intervals. The pharmacokinetics of rosiglitazone was not markedly affected by mild, moderate, or severe renal insufficiency. Slight increases (approximately 10%-20%) in mean unbound AUC0-infinity values were observed for each insufficiency group compared to the normal group but were not considered to be clinically relevant. Patients with severe insufficiency exhibited a 38% increase in mean fraction unbound, leading to an increase in total clearance, which resulted in a 19% to 24% lower mean total AUC0-infinity and Cmax values relative to the normal group. The rates of mild or moderate adverse events were similar for all groups; there were no severe adverse events. Impaired renal function does not markedly alter the pharmacokinetics of total or unbound rosiglitazone following a single dose of rosiglitazone. Therefore, the starting dose of rosiglitazone does not need to be adjusted in patients with renal impairment. Subsequent dose adjustments should be based on individual patient response.