Abstract An experiment was conducted to determine the effect of four tillage systems (moldboard plow, chisel plow, Paraplow and no-till) on soil aggregate shear strength and bulk density. Two soils, a Canisteo clay loam (fine-loamy, mixed (calcareous), mesic, Typic Haplaquoll) and a Haig silt loam (fine, montmorillonitic, mesic, Typic Argiaquoll) were used in this study. Soil samples were collected from the 0.075–0.15-m-depth increment in 1983 and the 0.075–0.15- and 0.225–0.30-m-depth increments in 1985. Shear strength of soil aggregates 0.02–0.03 m in diameter was measured by a fall-cone penetrometer and bulk density of the same aggregates was measured by gamma-ray attenuation. Aggregates were tested at soil water matric potentials ( ψ m) of −0.2, −1.1 and −4.0 kPa in 1983 and at ψ m of −0.2, −1.1, −4.0 and −7.9 kPa in 1985. Tillage for the 1983 growing season was conducted under very wet conditions, whereas tillage for the 1985 growing season was conducted under much drier conditions. Samples collected in 1983 showed little tillage effect on shear strength or bulk density. In 1985, tillage had an effect on shear strength and bulk density for the Haig soil, but not for the Canisteo soil. Much of the tillage effect on soil aggregate shear strength could be explained by tillage-induced changes in the aggregate bulk density. As bulk density decreased, soil aggregate shear strength decreased. Sampling depth had no effect on soil aggregate shear strength or bulk density. Matric potential had an effect on soil aggregate shear strength and bulk density. As matric potential decreased, both shear strength and bulk density increased.