The pawpaw [Asimina triloba (L.) Dunal] has great potential as a new fruit crop. A pawpaw variety trial was established in Fall 1995 in Princeton, Ky. as a joint Kentucky State Univ.-Univ. of Kentucky research effort with the objective to identify superior varieties for Kentucky. A randomized block experimental design was used with 8 replicates of 28 grafted scion selections on seedling rootstock. Cultivars being tested included Middletown, Mitchell, NC-1, Overleese, PA-Golden, Rappahannock, Shenandoah, Sunflower, Susquehanna, Taylor, Tay-two, Wells, and Wilson. The other 15 clones were selections from the PawPaw Foundation. In 2002 and 2003, the following parameters were examined: tree survival, trunk cross-sectional area (TCSA), average fruit weight, total fruit harvested per tree, average fruit per cluster, total yield per tree, and yield efficiency. In 2003, 54% of the trees had survived, with `Susquehanna' (13%) showing the poorest survival. Based on TCSA, most selections displayed excellent vigor, with the exception of the selections: 5-5 and `Overleese'. Average fruit weight was greatest in 1-7-2 (194 g), 1-68 (167g), 4-2 (321 g), 5-5 (225 g), 7-90 (166g), 9-58 (176 g), 10-35 (167 g), NC-1 (180 g), `Sunflower' (204 g), and `Shenandoah' (168g), with the smallest fruit in `Middletown' (70 g), `Wells' (78 g), and `Wilson' (88 g). The selections `Wilson' (81), `Middletown' (75), and `Wells' (70) had the greatest average number of fruit per tree, whereas 4-2 (9), 5-5 (17) and 8-20 (15) the fewest. Yield efficiency and average fruit per cluster also varied greatly among selections. Several pawpaw selections in the trial show promise for production in Kentucky.