Information regarding the impact of soil surface condition on soil-atmosphere exchange of gases is limited. In this study, fluxes and soil air concentrations of CO2, CH4, and N2O were monitored for 17 months at three central Ohio sites, including a bare (vegetation-free) soil, a mulch (covered with decomposed and fresh straw) site, and a 68-year-old deciduous forest (litter and canopy cover). Fertilizer was not applied to any of the sites. Relationships between daily fluxes of CO2 and soil temperature were described by linear and exponential functions. At the bare site, CO2 emission reached a maximum at 25°C, beyond which there was apparent insensitivity of soil respiration to temperature. Annual fluxes of CO2 and N2O from the bare, mulch, and forest sites (3.2, 4.9, and 4.6 Mg CO2-C ha-1 and 1.1, 1.3, and 1.4 kg N2O-N ha-1, respectively) were not significantly different. The bare and mulch sites were net CH4 emitters, but the forest was significantly different as a net CH4 sink (-2.2 kg CH4- C ha-1 y-1) with daily uptake averaging -0.92 and -0.63 mg CH4-C m-2 during dry and wet periods, respectively. Increased soil air concentrations of CO2 and N2O in the 10-20-cm soil depth coincided with higher emission rates. A generally similar trend was observed at the bare and mulch sites with respect to CH4. However, at the forest site, increased CH4 concentration in the upper soil layers was accompanied by increased uptake (-3.5 mg CH4-C m-2 d-1) in the summer but a net and short-lived emission (+0.5 mg CH4-C m-2 d-1) during spring thaw. N2O emission followed rainfall distribution, and the largest N2O pulses consistently followed termination of dry periods by rainfall events. Our conclusion is that wet-dry cycles are a more important controller of N2O emission from unfertilized soils than either temperature or soil cover.