The future size of the terrestrial methane (CH4) sink of upland soils remains uncertain, along with potential feedbacks to global warming. Much of the uncertainty lies in our lack of knowledge about potential interactive effects of multiple simultaneous global environmental changes. Field CH4 fluxes and laboratory soil CH4 uptake capacities were measured five times during three consecutive years in a California annual grassland exposed to eight years of the full factorial combination of ambient and elevated levels of precipitation, temperature, atmospheric CO2 concentration, and N deposition. Across all sampling dates and treatments, increased precipitation caused a 61% reduction in field CH4 uptake, but there were two quantitatively important multi-factor interactions that modified this general response. The wetter precipitation regime was more likely to reduce CH4 uptake in global change scenarios that included either warming or nitrogen (but not both), and elevated CO2 later in the growing season. Warming alone also decreased CH4 uptake earlier in the growing season, which was partly explained by a decrease methanotrophic activity. There was no evidence that the response in field CH4 uptake to increased precipitation was caused by substrate limitation or a change in methanotrophy. We hypothesize that the precipitation effects reflect an increase in endogenous soil CH4 production. Despite the complexity of interactions we observed, the outcome agrees with results from single-factor experiments: an increased terrestrial CH4 sink appears less likely than a reduced one. However, atmospheric CH4 models likely need to incorporate nonadditive interactions, seasonal interactions, and interactions between methanotrophy and methanogenesis.