MATERIALWe have studied the case records of 468 patients who underwent mitral valvotomy, 363 female and 105 male, to discover the incidence and site of systemic embolism before, at, and after operation, and the relation of embolism to cardiac rhythm and to the presence of clot in the left atrium at operation.Two hundred and seventy-two patients were operated upon in the Southampton Chest Hospital by the late E. F. Chin, K. S. Mullard, and I. K. R. McMillan during the period 1953-63, and 196 patients were operated upon by K. S. Mullard at Harefield Hospital in the period 1950-60.Thirty-eight (8-3 %) patients of this series have had a second operation for restenosis of the mitral valve, and one had a third operation. The mortality was 6-8 %, 3 % from embolism and 3-8% from other causes.SYSTEMIC EMBOLISM Of 468 patients with mitral stenosis, 76 (16-3 %) had a history of systemic embolism before operation. Five (1 %) suffered embolism at operation and 33 (7%) after operation. Fifteen (3 2%) patients had embolism before and after operation. Ninety-nine (21-1 %) patients in this series developed 114 embolic episodes before, at, and after operation (T7able I).Embolism was the prime indication for operation in 58 of the 76 patients with a pre-operative history of embolism; thus it was the prime indication in 12-4% of the cases.There were 15 cases of post-operative embolism among the 76 patients who suffered pre-operative embolism; it is not known how many of these 15 occurred in the 58 patients in whom embolism was the prime or only indication for valvotomy, but it is unlikely that all the 15 post-operative cases arose in the 58; the proportionate figure is 8 to 9, and, if so, there would seem to be a 12 to 15% risk of post-operative embolism, if valvotomy is advised where embolism is the prime symptomatic indication.The time of occurrence of post-operative embolism in 33 patients was as follows. In 13 patients it occurred hours after operation, in 14 after days, in two after weeks, and in four after months.Of the 114 embolic episodes, 74 (65%) were cerebral and 40 (35%) were peripheral. We tried to find out if there was any evidence that systemic embolism occurred more commonly on the left than on the right side. There were 32 cases of right hemiplegia, 24 developed left hemiplegia, and in 18 it was impossible to discover from the records whether it was right-sided or left-sided. Of 40 peripheral episodes, 16 were in the right leg or arm and 14 were in the left leg or arm.