Objective: To determine if ultrasonography is necessary for the diagnosis of pyloric hypertrophy in children and if in (clinical) doubt the ultrasonographic diagnosis of 'pylori hypertrophy' may constitute the indication for operation., Setting: Leiden University Hospital and Amsterdam Centre for Paediatric Surgery., Design: Retrospective study., Method: In the period 1985-1990, 83 children younger than 4 months were operated on for pyloric hypertrophy (group I; 73 of them had been examined by ultrasonography), while 15 underwent sonography because of suspicion of pyloric hypertrophy (group II), without operation. An attempt was made to find clinical and laboratory data that might be an indication of pyloric hypertrophy, and render supplementary ultrasonography unnecessary., Results: In group I, the combination of projectile vomiting, discernible gastric peristalsis after test meal and metabolic alkalosis was present in 34%, while a tumour in the upper abdomen was palpated in 25%; neither finding was present in group II. The sensitivity of one-time ultrasonography was 74%, that including follow-up examination in case of doubt, 82%. The predictive value of an ultrasonographic result positive for 'pyloric hypertrophy' was 98%, that of a negative finding 52%, the accuracy was 84%., Conclusion: In children younger than 4 months with the combination of projectile vomiting, gastric peristalsis after test meal and metabolic alkalosis, or with a palpable swelling in the upper abdomen, no ultrasonography is necessary to make the diagnosis of 'pyloric hypertrophy'. In case of clinical doubt, the indication for operation may be based on the result of ultrasonography.