The efficacy of colchicine combined with ursodeoxycholic acid (UDCA) and UDCA alone in the treatment of patients with nonadvanced primary biliary cirrhosis (PBC) was evaluated in a 2-year controlled study. Seventy-four patients with PBC who had been treated previously with UDCA (at least 8 months) but still had abnormal liver test results, especially elevated alkaline phosphatase activity, were randomized to be administered colchicine (1 mg/d, 5 days per week) (n = 37) or a placebo (n = 37). In addition, the patients were treated with UDCA (13-15 mg · kg-1 · day-1). The patients underwent clinical examination and liver tests every 6 months and upper endoscopy and liver biopsy at entry and at 2 years. Procollagen type III aminoterminal peptide (PIIINP), hyaluronic acid, and sulfobromophthalein (BSP) elimination kinetics were determined at entry and after 2 years. After 2 years of treatment, relative to UDCA, colchicine combined with UDCA did not significantly improve symptoms, laboratory findings (serum bilirubin level, alkaline phosphatase and alanine transaminase [ALT] activities, immunoglobulin [Ig] M level), serum markers of fibrosis, or histological features, except lobular inflammation. Colchicine did tend to slightly reduce the progression of esophageal varices; however, the difference was not significant. BSP elimination kinetics (45-minute retention percentage) was significantly improved when treated with colchicine. During the 2-year study, the only clinical complications were variceal bleeding in one patient administered colchicine and two administered the placebo. Two patients died from nonliver causes. One severe adverse effect (peripheral neuromyopathy) was observed in a colchicine-treated patient. In conclusion, this study suggests that colchicine appears to provide a slight advantage relative to UDCA alone in patients with nonadvanced PBC.