A small double-blind controlled trial to evaluate the short-term effects of D-penicillamine therapy was carried out in 24 patients with primary biliary cirrhosis (PBC). The daily dose of D-penicillamine was increased monthly by 250 mg until a total of 1 g daily was reached. Two out of 11 patients (18%) were withdrawn because of side-effects, as also were 4 out of 13 (31%) patients receiving the placebo. Transient improvement in symptoms was observed in 4 of 11 patients on D-penicillamine, but also in 5 of 13 patients from the placebo group. The proportion of patients showing a fall in serum IgM, IgG and hepatic copper was significantly larger for the D-penicillamine group than for the placebo group. No improvement in liver tests was observed, but the progression of inflammatory periportal liver cell destruction (piecemeal necrosis) was retarded in patients on D-penicillamine (p = 0.02). Data analysis within the D-penicillamine group showed that lowering the dose of D-penicillamine to 500 mg daily abolished the effect on the serum immunoglobulins and hepatic copper. The beneficial effect of D-penicillamine therapy appears to be small and dose-related; side effects should not prevent its use, provided the drug is introduced slowly.