To determine factors related to dehydration from diarrhoea, we conducted a hospital-based, case-control study in children aged 24 months or younger who had acute watery diarrhoea and attended Chonburi Regional Hospital in central Thailand during November 1988 through May 1989. The study compared 48 cases who had moderate or severe dehydration with 48 controls who had no dehydration. Both cases and controls belonged to low socioeconomic families and were living in urban slum areas. They had adequate health care facilities and access to ORS packets. Overall, 56% of the mothers used ORS solution at home. None of the mothers knew how to administer ORS, i.e. the fluid was not given at the onset of diarrhoea to prevent dehydration, and they gave no more than 60 ml over a 24-hour period to their dehydrated children. They also did not use home fluids. Multivariate analysis of data showed two factors significantly associated with dehydration: children's dirty fingernails that indicated inadequate maternal hygiene-related behaviour (Odds Ratio 6.4; 95% Confidence Intervals 1.5-27.6, p < 0.01), and frequency of vomiting in the 24 hours before rehydration (Odds Ratio 1.3; 95% Confidence Intervals 1.1-1.6, p < 0.001). Cases and controls had similar aetiologic agents and nutritional status. Providing proper education to mothers about oral rehydration therapy with special emphasis on the volume of ORS to be given, along with guidance to improve their personal hygiene should be considered important interventions in reducing the risk of dehydration and deaths from diarrhoea in these children.