The thymus is thought to be critical for T lymphocyte development, but is routinely (although not completely) removed to allow access to the heart in infants undergoing sternotomy. Three groups of children were studied. Group 1(n=40) underwent neonatal cardiopulmonary bypass and partial thymectomy(mean(se) age 12(1.2)days). Group 2 (n=40), also had bypass but no thymectomy(mean(se) age 48.4 (6.4) months). Group 3 (n=40) had thoracotomy, but no bypass or thymectomy (mean(se) age 24 (3.8) days). There was no significant difference in the total lymphocyte count preoperatively between the three groups. In all groups there was a fall in the total lymphocyte count post-operatively. The fall was significantly greater (p<0.01 ANOVA), and took longer to recover in group 1 (p<0.01 ANOVA). Chest radiographs of children in group 1 showed absence of thymic shadow post-operatively, but the return of a thymic shadow by two months in 20% of cases. We observed the lymphocyte count to return to normal before the thymus reached its pre-operative size. In four group 1 infants T-lymphocyte subsets (CD4/8, CD45RA/RO) were documented following thymectomy using flow cytometry, and antibody levels to routine tetanus toxoid vaccination were measured and were within normal limits. In one child surgery was complicated by a leak of thoracic duct fluid for six weeks. For three weeks the lymphocyte count was low. The thymic shadow in this infant returned to its preoperative size (as estimated by X ray and ultrasound) within five weeks, and by seven weeks was larger than it had been previously. The thymus in this case subsequently remained enlarged. Thymectomy during cardiac surgery is therefore unlikely to be followed by marked immunological sequelae. The thymus is capable of dramatic regeneration, and its size may be regulated to some degree by the rate of peripheral lymphocyte consumption or turnover.