OBJECTIVE The traditional method of aortic valve replacement with a homograft has been free-hand insertion in the subcoronary position. Recently, total root replacement has become increasingly popular. We present our experiences with both methods in this study. METHODS Between January 1, 1987 and March 31, 1996, 208 patients underwent homograft replacement of the aortic valve (free-hand subcoronary technique, n = 147 and root replacement, n = 61). The age of the patients ranged between 1.5 and 78 years with a mean age of 41 years. There were 55 females and 153 males. Ninety-four patients had infected aortic root (with 47 ring abscesses) and 114 patients had sterile aortic roots. In these series, patients with small aortic root, complicated endocarditis, dilated aortic annulus and aneurysm received aortic root replacement. RESULTS The hospital mortality of patients with non-infected and infected roots was 2.6% and 8.5% respectively making an overall hospital mortality of 5.2%. In patients with free-hand subcoronary valve implantation (AVR) and root replacement (ARR) techniques the hospital mortality was 3.4% and 9.8% respectively. The major risk factor for death was New York Heart Association functional class IV with sepsis. Eight years patient survival in patients with AVR and ARR was 95% +/- 2% and 86% +/- 4% respectively. Freedom from structural deterioration in patients with AVR and ARR was 96.5% +/- 2% and 98% +/- 3% whereas freedom from reoperation was identical 94% +/- 2% and 93% +/- 4% respectively. However, the incidence of reoperation in patients under 40 years of age, particularly in children (< 16 years of age) during the eight years' follow up was 27%. The rate of recurrent endocarditis was 3.6% in ARR patients and development of postoperative pseudoaneurysm formation occurred in 1.4% of AVR patients with endocarditis. CONCLUSIONS In conclusion, ARR technique provides low rate of reoperation in the early postoperative period. Cryopreserved homografts in the subcoronary position in adult patients < 40 years of age showed excellent medium term durability and hemodynamic performance providing evidence that long term outcome with a meticulous subcoronary implantation technique is comparable to that with the ARR technique.