The aim of this study was to define guidelines for intravenous contrast administration in cranial CT, as currently there are no recent guidelines based on a large series of patients. In 1900 consecutive patients (1480 adults and 420 children) pre- and post-contrast scan was analysed in order to assess the contribution of contrast enhancement to the diagnosis. The findings were grouped according to whether abnormalities were seen on the pre- and/or post-contrast scan, or whether no abnormalities were seen at all. Sensitivity, specificity, positive predictive value, negative predictive value and accuracy of a pre-contrast scan were used to determine validity. Intravenous contrast enhancement only contributes to the diagnosis if a suspicious abnormality is seen on the unenhanced scan or in the appropriate clinical setting (33.6%). In the remaining patients (65.6%) there is no diagnostic contribution, except for a small number of abnormalities (0.8%). These are often anatomical variants and have no therapeutic impact. The number of contrast-enhanced cranial CT examinations can significantly be reduced by using four general guidelines for contrast administration resulting in considerable cost savings without affecting the quality of service to the patient. These guidelines are defined by the clinical findings/presentation or by the findings on the unenhanced scan. The number of contrast-related complications will be reduced, which may have medicolegal implications. These guidelines can be applied in any radiology department.