Fenofibrate therapy is commonly prescribed to manage elevated triglyceride levels in diabetic patients but can cause a persistent increase in serum creatinine (SCr) levels in some patients within 3 months of starting therapy. The reversibility of this increase in SCr after cessation of fenofibrate therapy is uncertain. We conducted an on-drug/off-drug ancillary study to the ACCORD Lipid Trial to investigate changes in renal function after the Trial close-out visit. Eligible participants were recruited into a prospective nested 3 group study based on retrospective on-trial SCr levels: Fenofibrate (F-) Cases (¡Ý20% increase in SCr from baseline to month-4 visit, N=321); Fenofibrate (F-) Controls (¡Ü2 % increase, N=175); and Placebo (P-) Controls (no criterion; N=565). Serum creatinine and estimated glomerular filtration rate (eGFR) were measured at trial end and 6-8 weeks after discontinuation of fenofibrate or placebo therapy. At trial end, F-Cases had the highest SCr (±SE) (1.11±0.02 mg/dL) and lowest eGFR (±SE) (72.2±1.3 mL/min/1.73m2) vs. F-Controls (1.01±0.02; 80.0±1.7) and P-Controls (0.98±0.01; 83.3±1.0). After 51±10 (±SD) days off drug, F-Cases had a SCr (0.97±0.02) that was still higher (p=0.0008), and eGFR (83.5±1.3) still lower (p=0.004) than F-Controls (0.90±0.02; 90.0±1.8) but not different from P-Controls (0.99±0.01; 81.6±1.0). Five-year change in renal function for F-Controls compared to a matched subgroup of P-Controls (¡Ü 2% increase in SCr baseline-to-month 4, N=358) revealed F-Controls had significant SCr decline (-0.03±0.02 mg/dL, p=0.03) but no change in eGFR (0.5±1.4 mL/min/1.73m2); P-controls had significant SCr increase (0.05±0.01, p<0.0001) and eGFR decline (-6.5±1.0, p<0.0001). The differences between groups (F-Controls minus P-Controls) were also significant. In summary, at study end after 5.2 years of treatment, F-Cases returned to the same level of renal function as P-Controls, and F-Controls had net preservation of renal function compared to a matched subgroup of P-Controls. The fenofibrate-associated on-trial increases in SCr were reversible, and the reversal was essentially complete after 51 days off drug.