We describe here the broad spectrum of acute renal insufficiency occurring in the course of human immunoinsufficiency virus infection. In our renal unit in Tenon hospital, 90 human immunoinsufficiency virus-infected adult patients were admitted for acute renal insufficiency between June 1988 and December 1996. Sixty out of them had a pathological diagnosis. The remaining patients did not have renal biopsy because of obstructive renal failure (n = 2), bleeding risk (n = 11), or clinically evident hypovolemic and/or sepsis-related acute tubular necrosis (n = 17). Nine different causes of acute renal insufficiency were listed. Human immunoinsufficiency virus-associated nephropathy, the most specific human immunoinsufficiency virus-related renal disease, which was diagnosed in 14 patients, is characterized by focal and segmental glomerulosclerosis with an important hyperplasia and/or proliferation of podocytes and huge tubular distension. The rapid progression to end-stage renal failure was not a constant feature since 10/14 patients had a partial renal recovery. Hemolytic-uremic syndrome was the other major cause of acute renal failure in these patients (32 cases) and was found to be associated with active cytomegalovirus infection. Cytomegalovirus-infected cells were present in half of the renal biopsies performed in this group of patients. Furthermore, these patients had an increased plasma tissue-type plasminogen activator activity whereas its type 1 inhibitor was not significantly increased, as opposed to non human immunoinsufficiency virus-associated hemolytic-uremic syndrome. Half of the patients had a complete renal recovery. The other causes of acute renal insufficiency were 1) intratubular deposition of either drugs (Adiazine, Foscavir, Indinavir) in 13 patients, or monoclonal light chain in one patient with B cell-lymphoma; 2) lupus-like glomerulonephritis characterized in one case by a complete clinical remission after 6 month-treatment by antiproteases; 3) acute tubular necrosis. In this setting, rhabdomyolysis could reveal HIV infection. The heterogeneity of renal diseases could be explained by the variation of human immunoinsufficiency virus-associated infections along time and by the different drugs which permit a better survival. We can hypothesize that new HIV-associated diseases will occur with the long term use of antiproteases.