We compared the clinical, hemodynamic, and morphological findings in 18 young sheep in which porcine valvular bioprosthesis (eight animals) and bovine pericardial bioprostheses (10 animals) were implanted in the tricuspid position. At the time of terminal elective studies (mean of 5.2 +/- 0.2 months after implantation), six animals had ascites, 16 had hepatic congestion, and four had bioprosthetic valvular infection. Hemodynamic studies (n = 10) showed that the tricuspid transvalvular mean diastolic gradients were not different at implantation and at termination of the study (4.7 +/- 0.8 versus 4.9 +/- 0.9 mm Hg); however, tricuspid valve end-diastolic gradients increased from 1.2 +/- 0.6 to 3.9 +/- 0.5 mm Hg (p less than 0.01). Each of the 18 valves had calcific deposits. Quantitative studies revealed that implanted porcine valvular bioprosthesis (n = 7) contained a mean of 323 +/- 165 mg of calcium/gm of dry weight of cuspal tissue, in contrast to 0.2 mg/gm in unimplanted porcine valvular prostheses. Similarly, implanted bovine pericardial bioprostheses (n = 6) contained a mean of 421 +/- 115 mg of calcium/gm of dry weight of cuspal tissue, in contrast to 0.3 mg/gm im unimplanted bovine pericardial bioprostheses. Morphological findings in both types of bioprostheses included calcific deposits, collagen degeneration, leaflet immobilization and retraction, and fibrous sheaths. The latter were more extensive in bovine pericardial bioprostheses than in porcine valvular bioprostheses. We conclude: (1) that the pathological alterations which develop in bovine pericardial bioprostheses are generally similar to those in porcine valvular bioprostheses, but may be more severe; (2) that these alterations lead to physiological and clinical sequelae similar to those of bioprosthetic valvular failure in human subjects; and (3) that young sheep constitute an excellent experimental model for in vivo testing of bioprosthetic cardiac valves.