Switching from cyclosporine to tacrolimus without steroid pulse was suggested as a therapeutic option in chronic allograft nephropathy (CAN). Thirty-one renal transplant recipients with CAN were prospectively converted from cyclosporine to tacrolimus (group A), in parallel 31 matched cyclosporin A (CsA) patients (group B) without CAN were followed up for 30 months. In six matching patients of groups A and B inulin and para-aminohippurate (PAH)-clearances and mycophenolate were measured over a span of 3 months. Transplant biopsies of group A were scored according to BANFF. While group A presented with transplant dysfunction compared with group B before switching (2.7 +/- 0.16 mg/dl vs. 1.7 +/- 0.09 mg/dl; P < 0.001), transplant function was equal 30 months later: it ameliorated in group A (2.0 +/- 0.18 mg/dl vs. 2.7 +/- 0.16 mg/dl; P < 0.001) and decreased in group B (1.9 +/- 0.13 mg/dl vs. 1.7 +/- 0.09 mg/dl, P < 0.05). Especially, patients with biopsy scores I and II according to BANFF benefited from tacrolimus. Within 3 months, mycophenolate acid (MPA) levels increased under tacrolimus (P < 0.05) whereas inulin and PAH-clearances remained unchanged. At switching, antihypertensive treatment was more intense in group B, but this difference evened out. Adverse side effects were more frequent under tacrolimus. Patients with mild to moderate CAN significantly benefited from switching to tacrolimus. Increased MPA-levels under tacrolimus might have contributed to this effect.