During a consecutive 12-month period from January 1996 to January 1997 inclusive, 108 aortic valve replacements were performed by one group of surgeons in two community hospitals The majority of the valve replacements were done in combination with other procedures or were redo surgeries. Thirty-one patients had primary isolated aortic valve replacement. Fourteen patients underwent aortic valve replacement via a standard sternotomy, and seventeen patients underwent aortic valve replacement using a minimally invasive parasternal approach, as described by Dr. Cosgrove. There were no operative deaths in either group; however there was one hospital death in each of the two groups. Blood loss and postoperative pain were less in the minimally invasive group. Although the cross-clamp times were longer in the minimally invasive group, with a mean of 82.7 min as compared with 63.1 min in the standard group, the length of stay was shortened, with a median of 5 days in the minimally invasive group as compared to 7 days in the sternotomy group. In the follow-up which ranges from 4-15 months, all patients in the minimally invasive group were New York Heart Class I or II. Patients with the parasternal incisions are permitted to return to work much earlier than those with a standard sternotomy incision. The decreased blood loss and postoperative pain, combined with the anticipated ease of re-entry via a median sternotomy in the future (should redo aortic valve replacement become necessary), make this approach our procedure of choice in isolated primary aortic valve replacement.