BACKGROUND: Outcomes for a port-access (PA) approach for tricuspid valve operations have not been reported or compared to those using median sternotomy (MS). METHODS: Retrospective analysis was performed for 88 consecutive patients undergoing tricuspid valve repair or replacement using port-access techniques (n = 27, 1997-2000) versus sternotomy (n = 61, 1990-1997). PA procedures were performed through a 6 cm right fourth interspace thoracotomy. RESULTS: PA patients had lower ejection fractions (46% +/- 11% vs. 54% +/- 10%, p = 0.02), but had a similar incidence of previous surgery (17/27 (63%) vs. 33/61 (54%), p = 0.4). PA patients had more frequent concurrent mitral valve operations (22/27 (82%) vs. 37/61 (61%), p <0.05) and more tricuspid repairs versus replacement (24/27 (89%) vs. 29/61 (48%), p <0.01). PA patients had longer pump times (254 min. +/- 82 vs. 162 min. +/- 61, p = 0.001) but comparable clamp times (65 min. +/- 15 vs. 63 min. +/- 41, p = 0.9), lengths of stay (14 days +/- 14 vs. 16 days +/- 16, p = 0.6), mortality (2/27 (7%) vs. 9/61 (15%), p = 0.3), strokes (3/27 (11%) vs. 4/59 (7%), p = 0.9), and need for new pacemaker implantation (5/27 (19%) vs. 12/61 (20%), p = 0.9). CONCLUSIONS: PA provided excellent short-term results comparable to MS in relatively high-risk tricuspid valve patients. For tricuspid operations, PA may have the advantage of avoiding sternotomy or reoperative sternotomy at the expense of longer pump times.