Breath hydrogen tests were carried out on 157 children either because they had chronic diarrhea or because they were on disaccharide-free diets. Lactose malabsorption was common in patients with postgastroenteritis syndrome (43%), and sucrose malabsorption was readily detected in patients with congenital sucrase-isomaltase deficiency. Secondary sucrose malabsorption and small bowel bacterial overgrowth were also detected. In predicting clinical response to dietary change, the breath hydrogen test, as we perform it, was clearly the most specific and sensitive and had a predictive accuracy of 96%. Duodenal biopsy results obtained from 48 of the children gave a 23% incidence of misleading disaccharidase results (16.7% falsely normal, 6.3% falsely abnormal), but biopsy remains vital in the diagnosis of congenital sucrase-isomaltase deficiency. False negative breath hydrogen results were obtained on occasions (4%) but in most instances were related to recent antimicrobial therapy or failure of the breath test mechanics (e.g., vomiting, length of sampling).