We compared Hexabrix 320 (580 mOsm kg-1) with Conray 420 (2500 mOsm kg-1) for left ventriculography using a prospective randomised double-blind protocol. One hundred consecutive patients with suspected coronary disease were assigned to Hexabrix (52) or Conray (48) for left ventriculography (dose 10 ml m-2 BSA; flow rate 12 ml s-1). Thirteen patients found Hexabrix unpleasant compared with 24 receiving Conray; overall the feelings of warmth and discomfort were less with Hexabrix than Conray (p less than 0.01 and p less than 0.02 respectively). The incidence of nausea, vomiting, and hypersensitivity was similar. Angiographic quality was better with Conray than with Hexabrix (p less than 0.05). Average changes in heart rate and systolic pressure were similar, though there was greater variation in systolic pressure change after Conray (p less than 0.025). End diastolic pressure increased more after Conray than after Hexabrix (p less than 0.05). These slight advantages of Hexabrix over Conray may be valuable in patients requiring multiple angiograms or in those with impaired cardiac function, but do not justify its use for routine angiography.