A survey of 763 patients with rheumatoid arthritis and 145 with osteoarthritis in six clinics in New Zealand showed no association between aspirin intake and a score designed to detect analgesic nephropathy. Analgesic nephropathy was diagnosed clinically in three patients taking APC (aspirin, phenacetin, and caffeine or codeine or both) and in one who took aspirin and phenylbutazone and was suspected in one who took aspirin and paracetamol. Isolated aspirin was not implicated. The study showed that most people can take large quantities of salicylates without renal injury. The findings are, however, consistent with the view that there is a risk from APC compounds taken in large quantity, but the numbers at risk in this study were small. Aspirin may have an additive effect with other analgesics in causing renal damage. An increased frequency of urinary tract symptoms in those taking analgesics requires further investigation.