Diaphragm paralysis has been reported radiologically after cardiac surgery with an incidence ranging from 30% to 75% of patients. We studied 100 consecutive patients undergoing open heart operations, half of whom received ice/slush topical hypothermia (group 1) and half of whom did not (group 2). Chest radiology and diaphragm screening were performed at 1 week, 1 month, and every 6 months thereafter in all patients with an elevated diaphragm. Phrenic nerve conduction time was measured in all patients in whom there was radiological evidence of diaphragm paralysis 1 week postoperatively. The two groups were similar in terms of age and sex. Aortic cross-clamp time was less in group 1 (61.5 +/- 15.6 minutes) compared with group II (74.4 +/- 20.8 minutes), although this difference was not significant. Significant differences, however, were found for radiological evidence of partial left lower lobe collapse (82% in group 1 versus 32% in group 2; p less than 0.01) and for radiological evidence of diaphragm paralysis (32% in group 1 versus 2% in group 2; p less than 0.001) within the first postoperative week. Unilateral diaphragm paralysis developed in 16 group 1 patients (15 left sided, 1 right sided) compared with only 1 patient in group 2. In these 16 group 1 patients, diaphragm paralysis was still present in 12 (75%) at 1 month and in 5 (31.3%) at 1 year postoperatively. There were no significant differences between the two groups in terms of postoperative arrhythmias, myocardial infarction, or mortality. Phrenic nerve conduction time was found to be a sensitive indicator of phrenic nerve cold injury and recovery.(ABSTRACT TRUNCATED AT 250 WORDS)