A retrospective follow-up study in the Netherlands on the risk of fracture of Björk-Shiley convexo-concave valves concluded that prophylactic replacement is advisable for certain groups of patients. We have examined valves explanted from twenty-two patients because they met epidemiological criteria for reoperation, with or without the presence of moderate cardiac impairment, or because there were other cardiac complaints not related to the valve. No information was available before explanation to suggest a valve defect. All patients survived their operations. Of the twenty-four valves, seven (29% [95% CI 13-52]) had fracture of one of the legs of the outlet strut (single-leg strut fracture [SLF]). Two other valves had features that suggested fatigue defects. As in the previous study, 70 degrees valves had the highest risk of SLF (five of sixteen, two aortic and three mitral). However, two of eight 60 degrees valves (both aortic) also had SLF. Current hazard calculations and explanation recommendations may need to be revised. Since several of the defective valves were welded by the same person, knowledge of manufacturing details may help in estimation of fracture risk for an individual patient.