Between 1974 and 1990, 58 patients underwent tricuspid valve replacement with porcine bioprostheses (Hancock 42, Carpentier-Edwards 16) during multiple valve replacement (double, 21; triple, 37). Perioperative mortality was 12%; 16 patients died later, mostly from cardiac causes. Actuarial survival (1 patient lost to follow-up) was 81% +/- 11% at 5 years, and 60 +/- 17% at 10 years. Reoperation because of Hancock prosthesis deterioration was performed in 2 patients at 11 and 15 years, respectively. At last follow-up (mean 108 +/- 48 months), 82% of survivors (28/34) were functionally improved. Doppler echocardiography was performed in 29 patients in February 1991. In 21 patients, after 88 +/- 40 months of follow-up, the bioprosthesis was normal, there was no leaflet malformation, no significant tricuspid regurgitation and the mean diastolic transprosthetic gradient (DTPG) was 3.8 +/- 1.7 mmHg. In 7 patients (follow-up: 129 +/- 40 months, P less than 0.05), there was moderate dysfunction (all Hancock prostheses) with leaflet sclerosis, tricuspid regurgitation grade 2, and mean DTPG 5.7 +/- 1.8 mmHg (P less than 0.05). Only 1 patient (Hancock prosthesis implanted in 1981) had severe tricuspid prosthesis stenosis with very thickened leaflets and mean DTPG 13 mmHg. Pulmonary artery hypertension (most often fixed) was present in 11 patients, associated with a poor functional result and a significantly higher DTPG. We conclude that porcine bioprostheses in tricuspid position have an acceptable long-term durability and satisfactory performance. Prosthetic dysfunction correlates with the length of follow-up of patients and with the presence of fixed pulmonary artery hypertension.