The authors studied with duplex-Doppler US 28 renal transplant recipients in 31 clinically different episodes, during the early postoperative period. Morphological data were thus obtained, as well as hemodynamic information. According to the literature on the subject, a pulsatility index (PI) greater than 1.5 was considered as abnormal. US diagnosis was retrospectively compared with final clinical diagnosis and with response to therapy. In one case, the kidney was surgically removed. We evaluated US sensitivity and specificity in the diagnosis of acute rejection with real-time US, Doppler alone and combined with duplex. A PI greater than or equal to 1.5 corresponded to acute rejection, with 60% sensitivity and 85.7% specificity. With a PI greater than 1.8, sensitivity decreased to 50%, but specificity increased to 100%. The severest changes in Doppler waveforms had a bad prognostic significance. Besides poor specificity--which is so often emphasized in literature--our results chiefly demonstrated sensitivity limitations, partly corrigiable with real-time US signs, together with Doppler PI (sensitivity: 90%, specificity: 85.7%). Duplex-Doppler US, in spite of its well-known limitations, remains therefore a simple, rather reliable and non-invasive technique to study renal transplant complications.