In a survey of the literature, phenacetin appears to be responsible for an increased risk of urothelial tumours in humans, especially renal pelvis tumour. Few observations have suggested an association between bladder cancer and phenacetin containing analgesics however. To assess this association in France, a case-control study of bladder cancer (143 cases, 120 controls) was undertaken. We point at some methodological difficulties, such as difficulties of recall, definition of a control group, non-responses, possible French distinctive features of analgesics use. Results are not able to confirm the suspected relationship between bladder cancer and phenacetin use.