Anthropogenic N deposition has been well documented to cause substantial impacts on the chemical and biological properties of forest soils. In most studies, however, atmospheric N deposition has been simulated by directly adding N to the forest floor. Such studies thus ignored the potentially significant effect of some key processes occurring in forest canopy (i.e., nitrogen retention) and may therefore have incorrectly assessed the effects of N deposition on soils. Here, we conducted an experiment that included both understory addition of N (UAN) and canopy addition of N (CAN) in two contrasting forests (temperate deciduous forest vs. subtropical evergreen forest). The goal was to determine whether the effects on soil exchangeable cations and microbial biomass differed between CAN and UAN. We found that N addition reduced pH, BS (base saturation) and exchangeable Ca and increased exchangeable Al significantly only at the temperate JGS site, and reduced the biomass of most soil microbial groups only at the subtropical SMT site. Except for soil exchangeable Mn, however, effects on soil chemical properties and soil microbial community did not significantly differ between CAN and UAN. Although biotic and abiotic soil characteristics differ significantly and the responses of both soil exchangeable cations and microbial biomass were different between the two study sites, we found no significant interactive effects between study site and N treatment approach on almost all soil properties involved in this study. In addition, N addition rate (25 vs. 50<ce:hsp sp="0.25"/>kg<ce:hsp sp="0.25"/>N<ce:hsp sp="0.25"/>ha −<ce:hsp sp="0.10"/>1 <ce:hsp sp="0.25"/>yr −<ce:hsp sp="0.10"/>1 ) did not show different effects on soil properties under both N addition approaches. These findings did not support previous prediction which expected that, by bypassing canopy effects (i.e., canopy retention and foliage fertilization), understory addition of N would overestimate the effects of N deposition on forest soil properties, at least for short time scale.