With the aim to asses the thrombogenicity of "treated" bovine pericardial tissue, the incidence of thromboembolism with Ionescu-Shiley pericardial bioprostheses was analyzed in 349 operative survivors since January 1977. A total of 376 valves were available for the study. Mean age was 59 years. Preoperatively atrial fibrillation was present in 68.2% (91.7% of the mitral replacement group) and 62% were in NYHA functional class III-IV; in 10.6% previous embolism had been documented. Early anticoagulation (8-12 weeks) was administered in the first 232 patients (113 mitral replacements, 98 aortic replacements and 21 multiple replacements) and the rest of the patient population had no anticoagulation at all. Thromboembolic events were documented in 13 patients (3.72%) or 0.84% per patient/annum (1.49% per patient/annum for mitral and multiple replacements and 0.13% per patient/annum for aortic replacements). Early thromboembolism was detected in 6 patients (3 months after surgery), all in anticoagulated patients, three being in sinus rhythm. No valve thrombosis was detected in this series. Three patients with late thromboembolism were in NYHA functional class III-IV. Ten thromboembolic events were central (1 died and three had permanent neuro-residua). In two the location was peripheral and another one in the kidney. The Ionescu-Shiley pericardial bioprosthesis after 11 years has demonstrated a low thrombogenicity in the absence of chronic anticoagulation. Early anticoagulation does not influence the rate of thromboembolism during this period and therefore we believe that this therapeutic approach is not justified despite the presence of risk factors.