Background: Acute rejection episodes after renal transplantation are an important clinical challenge, despite use of multidrug immunosuppressive regimens. We did a prospective, multicentre, randomised, double-blind trial to investigate the impact of the addition of sirolimus, compared with azathioprine, to a ciclosporin and prednisone regimen. Methods: 719 recipients of primary HLA-mismatched cadaveric or living-donor renal allografts who displayed initial graft function were randomly assigned. after transplantation, sirolimus 2 mg daily (n = 284) or 5 mg daily (n = 274), or azathioprine (n = 161). We assessed the primary composite endpoint of efficacy failure, occurrence of biopsy-confirmed acute rejection episodes, graft loss, or death, and various secondary endpoints that characterise these episodes at 6 months and 12 months. Analyses were done by intention to treat. Findings: The rate of efficacy failure at 6 months was lower in the two sirolimus groups (2 mg 18.7%, p = 0.002; 5 mg 16.8%, p < 0.001) than in the azathioprine group (32.3%). The frequency of biopsy-confirmed acute rejection episodes was also lower (2 mg 16.9%, p = 0.002; 5 mg 12.0%. p < 0.001; azathioprine 29.8%). At 12 months, survival was similar in all groups for grafts (97.2%, 96.0%, and 98.1%) and patients (94.7%, 92.7%, and 93.8%). Patients on sirolimus showed a delay in the time to first acute rejection episode and decreased frequency of moderate and severe histological grades of rejection episodes and related antibody treatment, compared with the azathioprine group. Rates of infection and malignant disorders were similar in all groups. Interpretation: Use of sirolimus reduced occurrence and severity of biopsy-confirmed acute rejection episodes with no increase in complications. Further studies are needed to establish the optimum doses for the combined regimen.