BACKGROUND: Cyclosporine represented a major advance in the medical management of patients with organ transplantation, but its use is limited by the frequent occurrence of hypertension and renal toxicity diagnosed by invasive renal biopsy. Renal histology shows a specific arteriolopathy. It was hypothesized that cyclosporine may also induce subclinical microvascular changes in the skin that might be detected noninvasively by a combination of dynamic capillaroscopy [capillary blood cell velocity (CBV)] with and without intravenous Na-fluorescein (NaF) injection and laser Doppler fluxmetry (LDF). METHODS: The nailfold skin microcirculation was evaluated in 112 consecutive renal transplant recipients (54 +/- 11 years old; 70 males and 42 females) receiving cyclosporine. The investigation was made the same day as a routine renal biopsy performed in all patients more than two years after transplantation. Renal biopsies were blindly classified as positive (N = 33) when significant specific signs of cyclosporine toxicity were clearly observed (AH2-AH3) and were otherwise negative (AH0-AH1, N = 79) according to the Banff classification. RESULTS: Time to fluorescence peak after NaF injection (tpNaF) was significantly longer in patients with positive biopsies than in patients with negative biopsies (13.9 +/- 8.1 vs. 17.5 +/- 9.4 sec, P = 0.009). All patients but three with negative biopsies (93%) had a tpNaF less than 10 seconds (sensitivity 91%, negative predictive value 93%). On the other hand, CBV, LDF, plasma levels of cyclosporine, and endothelin were similar in the two groups. CONCLUSION: Nailfold fluorescence capillaroscopy is an accurate and simple mean to rule out cyclosporine toxicity in renal transplant recipients. A normal test could avoid invasive renal biopsy in about 40% of the patients. Renal biopsy would, however, still be indicated when the test is abnormal