Primary hyperoxaluria (PH) is a heterogeneous disease with a variable age of onset and a variable progression into kidney failure. Early diagnosis is mandatory to avoid the damaging effects of systemic calcium oxalate deposition. In 1997, we initiated a nationwide survey of American nephrologists to ascertain epidemiological data and current practices. PH was reported in only 102 patients, with PH I in 79 and PH II in 9; 14 patients were not classified. Most patients were Caucasian (84%). Main symptoms at diagnosis were urolithiasis (54.4%) and nephrocalcinosis (30%). A significant delay of diagnosis was seen in 42% of patients and 30% of patients were diagnosed only at end-stage renal disease (ESRD). Diagnosis was usually based on history and urinary oxalate excretion. Glycolate and l-glyceric acid excretion were rarely determined. To determine the enzyme defect, a liver biopsy was performed in 40%. Even at ESRD, only 56% of patients received an adequate diagnostic work-up. Half of the patients showed 'good' or 'fair' pyridoxine sensitivity. In addition to B(6), most patients received either citrate or orthophosphate. Kidney transplantation (KTx) failed in 19 of 32 transplants ( n=27 patients) and was due to recurrent oxalosis in 8 transplants. Liver Tx was performed after KTx in 5 patients (1 patient died). Combined liver-kidney Tx in 21 patients (in 9 patients after failure of KTx) achieved good organ function in 13 patients; 7 patients, however, died shortly after transplantation. In conclusion, the time between first symptom and diagnosis of PH must be minimized, and the diagnostic procedures have to be improved. The cases of unclassified hyperoxaluria suggest the possibility of additional type(s) of PH. As isolated KTx failed in 59% of patients, combined liver-kidney Tx seems to be the better choice in place of isolated KTx as the primary transplant procedure.