A study group of 623 working women 30 to 49 years old with objective evidence of intake of phenacetin-containing analgesics and a matched control group of 621 women without such intake in 1968 were assessed six times from 1969 to 1978 for laboratory evidence of urinary-tract disorders. The two groups did not differ in development of bacteriuria, hematuria, or proteinuria. However, a low specific gravity of urine (study group vs. control group, 23 vs. 7 per cent) and a raised level of serum creatinine (6.7 vs. 0.9 per cent) were significantly more frequent in the study group (P less than 0.001). Adjusted analyses of mortality over 11 years showed significant differences between the groups in overall mortality (study group vs. control group, 39 vs. 13 deaths; P less than 0.001), mortality due to urinary-tract disorders (P = 0.033), and cardiovascular diseases (P = 0.008). We conclude that heavy users of analgesic mixtures in the course of a decade have a higher incidence of both abnormal kidney function and kidney-related mortality than do casual users and nonusers, but the absolute incidences remain small even among heavy users.