Six patients, five of whom had normal and one impaired renal function, and all suffering from purulent arthritis caused by cephalosporin-sensitive germs, were given a seven-day course of 8 g cephacetrile daily. On the first day, 6 g were administered by continuous intravenous infusion at the rate of 500 mg/h, followed by 2 g over a further 45 min. On days 2 to 7, the patients received 2 short infusions of 4 g each at an interval of 12 h. In four patients with normal renal function, serum half-life ranged from 0.8 to 1.4 h, serum levels during continuous infusion from 19 to 31 microgram/ml, and total clearances from 265 to 434 ml/min. In one patients, these values were 1.6 h, 70 microgram/ml and 131 ml/min respectively (small volume of distribution). The concentrations in the synovial fluid varied from 2 to 29 mcirogram/ml; they were generally lower than the serum levels, but clearly exceeded the minimum inhibitory concentrations for germs commonly present in purulent arthritis. In five patients, the synovial fluid became germ-free and the arthritis was clinically cured. In the case presenting with renal insufficiency, the serum half-life was 5.8 h. During continuous administration, a steady state was not attained; peak serum levels amo9nted to 75 microgram/ml and the total clearance to 61 ml/min. The cephacetrile concentrations in the synovial fluid were very high (26 and 67 microgram/ml). In this case, in which the renal insufficiency associated with mycosis fungoides was present before the treatment, renal function deteriorated futher during treatment while the arthritis improved.