OBJECTIVE: Whether chest radiographs (CXRs) in mechanically ventilated patients should be routinely obtained or only when an abnormality is anticipated remains debated. We aimed to compare the diagnostic, therapeutic and outcome efficacy of a restrictive prescription of CXRs with that of a routine prescription, focusing on delayed diagnoses and treatments potentially related to the restrictive prescription. DESIGN: Randomized controlled trial. SETTING: Intensive care unit of the Avicenne Teaching Hospital, Bobigny, France. PATIENTS AND PARTICIPANTS: All consecutive patients mechanically ventilated for > or = 48h between January and June 2006. INTERVENTIONS: Patients were randomly assigned to have daily routine CXRs (routine prescription group) or clinically indicated CXRs (restrictive prescription group). MEASUREMENTS AND RESULTS: For each CXR, a questionnaire was completed addressing the reason for the CXR, the new findings, and any subsequent therapeutic intervention. The endpoints were the rates of new findings, the rates of new findings that prompted therapeutic intervention, the rate of delayed diagnoses, and mortality. Eighty-four patients were included in the routine prescription group and 81 in the restrictive prescription group. The rates of new findings and the rates of new findings that prompted therapeutic intervention in the restrictive prescription group and in the routine prescription group were 66% vs. 7.2% (p < 0.0001), and 56.4% vs. 5.5% (p < 0.0001) respectively. The rate of delayed diagnoses in the restrictive prescription group was 0.7%. Mortality was similar. CONCLUSIONS: Restrictive use of CXRs in mechanically ventilated patients was associated with better diagnostic and therapeutic efficacies without impairing outcome.