We performed a randomised, double-blind, prospective trial to discover whether intravenous ketorolac 10 mg made up to 2 ml with saline, with or without venous occlusion for 2 min, reduces the pain on injection of propofol. In 90 patients, pain scores were obtained during injection of propofol following pretreatment of the vein with saline, ketorolac or ketorolac with venous occlusion. Pain on injection of ketorolac was more common than with saline (p = 0.02). The incidence of severe pain following propofol was reduced by ketorolac with venous occlusion (p = 0.019) compared with saline or ketorolac without venous occlusion. There was no difference in venous sequelae at 7 days postoperatively between the groups. Our results suggest that pain on injection of propofol may be related to release of local kininogens and that nonsteroidal anti-inflammatory drugs may have a role in reducing that pain.