We conducted soil moisture manipulation experiments in a red pine (PinusresinosaAit.) plantation at the Harvard Forest (Petersham, Mass.) in August 1992 and May 1993. To manipulate soil moisture, we added 10 cm of groundwater to 1-m2plots and allowed the soils to dry down to their pretreatment moisture contents. We measured methane (CH4) flux, soil moisture, and temperature prior to and after the water addition. Soils in both the control and watered plots were usually sinks for atmospheric CH4. Average consumption rates by control soils ranged from 0.12 to 0.17 mg CH4-C•m−2•h−1. Methane consumption rates by watered soils ranged from 0 to 0.12 mg CH4-C•m−2•h−1and were inversely related to the moisture content of the upper 10 cm of mineral soil. Linear regression between soil moisture and CH4consumption explained 78% of the variability (CH4consumption = 0.001 75 (percent water filled pore space)–0.1957). Using this empirical relationship, we predicted CH4consumption by soils at three other locations in the Harvard Forest, which agreed closely (r2 = 0.7574) with rates measured in the spring, summer, and fall of 1988–1992. Results from our study suggest that soil moisture is a good predictor of methane uptake by these forest soils and may be used to predict how future changes in soil moisture resulting from alterations in regional precipitation patterns will affect the strength of this terrestrial CH4sink.