In a prospective, randomized, double-blind study, 49 patients underwent lumbar myelography using iotrol (24 patients) or metrizamide (25 patients). The diagnostic imaging adequacy of iotrol was comparable with that of metrizamide. After iotrol myelography, adverse reactions were fewer, less severe, and of shorter duration than were those following metrizamide myelography. Thirteen of 24 patients (54%) receiving iotrol reported some adverse reactions compared with 24 of 25 patients (96%) receiving metrizamide. Five moderate and one severe adverse reaction occurred in the group receiving iotrol. Fourteen moderate and eight severe adverse reactions occurred in the group receiving metrizamide. Thirty-eight patients underwent electroencephalography both before and after myelography (19 iotrol and 19 metrizamide). None of the EEGs obtained after iotrol myelography changed from baseline, while seven of the EEGs obtained after metrizamide myelography showed changes from baseline. Iotrol was judged superior to metrizamide as a contrast medium in this patient population.