In three trials, 2004 crop loads were adjusted at the balloon stage of blossom development on heavily cropped whole apple trees by clipping all flower pedicels within a cluster while leaving the spur intact. Trees were adjusted to 0% crop (all flowers removed), 50% crop (half of all clusters removed), or left as 100% crop (no flowers removed). On `Cameo'/Bud.9, 400 ppm GA4+7 were applied to trees of each crop level at petal fall, 10 mm, 20 mm, or left unsprayed. At each crop load, GA4+7 marginally diminished the 2005 return bloom regardless of application timing, but the 2004 crop level was far more influential in 2005 flowering. In a second `Cameo'/Bud.9 trial, 0, 300, 600, or 900 ppm ethephon were applied to whole trees of each crop level at 45 DAFB. Ethephon generally demonstrated a rate response in improving the 2005 return bloom, but the 2005 flowering was more dramatically influenced by 2004 crop levels. On `Honeycrisp'/M.9, 300 ppm GA4+7 were applied to whole trees of each crop level at 10 mm. GA4+7 diminished 2005 return bloom at the 50% crop load, but spray effects were not as clear at the extreme 2004 crop levels. These results suggest that commercial floral inhibitors and promoters have difficulty overcoming endogenous effects of heavy or light bloom and crop in severely alternating apple trees. In a fourth trial, lightly cropped organic `Fuji'/MM.106 trees were sprayed with 0, 150, 300, or 450 ppm GA4 at petal fall, 10 mm, or 20 mm timings in 2004. The 2005 return bloom was inversely correlated with spray rates, with 10 mm showing more floral inhibition than other timings. Overall, `Cameo' was less sensitive to GA and ethephon than `Honeycrisp' or `Fuji'.