Urinary incontinence occurring at least once weekly was found in 32% of residents in 30 Local Authority Residential Homes for the Elderly. Incontinence was significantly commoner among female residents. One hundred and seventy-four incontinent residents were randomly selected in order to assess the effect of medical intervention (study group). A further 104 incontinent residents were selected to act as controls (control group). The vast majority of the study group showed evidence of chronic brain failure and 40% had impaired mobility. Almost half had been incontinent since admission. Urodynamic studies were required for management in only two (1.1%). Most (93%) were managed in accordance with a clinical diagnosis of an unstable bladder. There was a reduction in daytime incontinence in 40% of the study group, but this was matched by a reduction in 29% of the control group and the difference was not significant. The reduction in nocturnal incontinence in 41% of the treatment group was significantly different from 23% of the control group (P = 0.016).