Treatment of primary nocturnal enuresis using DDAVP is based upon the hypothesis that antidiuretic hormone (ADH) secretion is insufficient at night. The known efficacy of the treatment on the one hand, and persisting doubts about its theoretical basis on the other, formed the background of the present study. Ten children (mean age 10.5 years) with primary nocturnal enuresis were compared with a corresponding control group of eight patients. Diurnal and nocturnal urine production, ADH secretion, and plasma osmolality were determined. No differences between the two groups were found for urine production, ADH levels during day and night, or plasma osmolality. However, in order to regulate plasma osmolality the enuretic children required a markedly greater output of ADH: 2.87 (95% confidence interval 0.091 to 40.35) pg/ml/mmol/kg v 0.56 (0.08 to 1.03) in the controls (p < 0.01). The results are consistent with the established fact that ADH secretion is a function of plasma osmolality, and they contradict the hypothesis that urine production is increased at night in enuretics because of lower ADH secretion. The findings do not solve the uncertainties in the pathogenesis of enuresis but they suggest there might be a difference between enuretic children and controls at the ADH receptor level.