In a prospective clinical study of the outcome of alarm treatment in nocturnal enuretics, 60 children were included: 40 boys and 20 girls, mean age 8.2 years (range 5.1-14.4). All were treated with enuresis alarms and had 2 or more enuretic events during the initial 14 days of treatment. None had diurnal enuresis. In each child, the enuretic and voluntary voiding frequencies during the initial 14 and last 14 days of treatment were compared. We found that 43 children had a 75% reduction or more of the enuretic events. 28 children substituted the former enuretic events by sleep, 15 changed the enuresis by voluntary voidings. Only 17 children had no effect of the alarm treatment. No parameters were found to predict the outcome. In conclusion, the outcome of successful alarm treatment occurs in two distinct forms. Either the child is left asleep without wetting his bed; or the child wakes up spontaneously from sleep and goes to the bathroom.