A random sample of 6,000 women from the birth cohorts 1900-1920 were invited to participate in an investigation of urinary incontinence by completing and returning an enclosed questionnaire. Of the 4,206 women who returned the completed questionnaire, 677 (16.9%) complained of urinary incontinence and accepted an invitation to be examined and treated at our clinic. The first 150 consecutive patients recruited via this questionnaire who attended the clinic underwent a detailed neurological examination. The prevalence of neurological signs amongst this group of women was low, only 2% were considered to have a focal encephalopathy, 1% had an organic brain syndrome, 2% had a myelopathy and a further 1% were considered to have a polyneuropathy. There was no significant difference in the prevalence of neurological signs between the group of women with urinary incontinence and an age-matched control group from the total population. Thus, urinary incontinence in an unselected population of elderly women appears to be mainly dependent on other aetiological factors rather than neurological dysfunction.