The records of all (5080) patients presenting to a district general hospital with acute abdominal pain over a 4-year period were examined. The contribution of abdominal radiographs to the assessment of patients with suspected appendicitis, urinary tract infection, and non-specific abdominal pain was evaluated, these conditions accounting for 48 per cent of patients with abdominal pain and 32 per cent of those with abdominal radiographs. Any positive information from these radiographs was less likely to be helpful than incidental or inconsistent (and hence potentially misleading). Because of this high 'false positive' rate it is suggested that if the initial diagnosis is suspected appendicitis, urinary tract infection, or non-specific abdominal pain, there is little value in the routine use of abdominal radiographs.