In 1997 and 1998, we determined the effects of defoliation on return bloom and fruit set following a light cropping year. In one study, `Braeburn' trees were hand-thinned to a crop density (CD) of 3 fruit/cm 2 trunk cross sectional area (TCSA) in late May 1997, and then either completely defoliated or half of the tree defoliated by hand on one of five dates between June and Sept. 1997. Compared to a nondefoliated control, both whole and half-tree defoliation on all dates reduced fruit count and yield efficiency (kilograms per square centimeter of TCSA) and affected fruit weight, starch, firmness, and soluble solids in 1997. In 1998, return bloom and fruit set were reduced by most 1997 defoliation treatments. Compared to other dates, defoliation on 3 July caused the greatest reduction in return bloom in both whole and half-defoliated trees. In another study, `Braeburn' trees were hand-thinned to a CD of 5 in late May 1998; complete defoliation by hand on 1, 15, or 29 July reduced return bloom and fruit set in 1999; the 1 July treatment resulted in zero return bloom. `Golden Delicious' and `York' trees were thinned to a CD of 3 in late May 1998 and were hand-defoliated on 21 July or 12 August by removing every other leaf or removing three of every four leaves over the entire tree. In 1999, return bloom and spur and lateral fruit set were reduced by all defoliation treatments. Fruit set was most reduced by the 12 Aug. treatment. Fruit set for `York' was lower than for `Golden Delicious' in all cases.