A prospective, open, multicentre, randomized study with a blinded radiological end-point was started in 1991. The aim of the study was to assess whether cyclosporin A (CyA) controls ongoing anatomical damage in active early rheumatoid arthritis (RA) better than conventional disease-modifying anti-rheumatic drugs (DMARDs) as used in everyday clinical practice. A total of 340 consenting patients with early RA (mean duration 1.4 yr) were recruited; 167 were randomized to CyA 3mg/kg per day and 173 to DMARDs. Hand, wrist and foot X-rays were blindly scored by a central committee of three radiologists using the Larsen-Dale method. Any side-effects were carefully recorded. The control of clinical symptoms was similar in both groups. Radiological evaluation of 284 patients (141 on CyA; 143 on DMARDs) after 12 months showed a significant decrease in the mean progression in the eroded joint count (1.3 +/- 3.1 vs 2.4 +/- 3.0, P < 0.001). There was also better maintenance on treatment with CyA than in the group treated with DMARDs (89.2 vs 77.5%, respectively; P = 0.002). CyA seems to offer greater control of ongoing anatomical joint damage in early RA than conventional DMARDs after 12 months.