335 of ventricular septal defect treated surgically in our hospital between June 1989 and November 1996 were studied retrospectively. 15 of these cases had a preliminary diagnosis of ruptured membraneous septal aneurysm, which however, during the operation it was found out that 6 of them did not have true membranous septal aneurysm; rather, had such an appereance due to the fact that tricuspid septal leaflets were adhered to the edges of large perimembraneous type of VSD’s, forming incomplete spontaneous closure. Having dissected along the annulus leaflet, the disclosed large defects were patched and leaflets subsequently repaired. There was patched and leaflets subsequently repaired. There was no morbidity or mortality either peri- or postoperatively. During the follow up period of 34.8 months there were no findings suggesting residual VSD or dysfunction of tricuspid valve. It may be difficult to differentiate perimembraneous VSD’s that are spontaneously closed by tricuspid septal leaflets and membraneous septal aneurysms. The anatomical assessment and surgical management of such cases present special consideration.