Multi-factor experiments greatly increase the conceptual and analytical complexity of manipulative global change projects, yet they are a necessary step in determining the potential for interactive effects of various factors of global change on both above and belowground response variables. The Old-Field Community Climate and Atmosphere Manipulation (OCCAM) experiment was designed to assess interactive effects of [CO₂], warming, and soil moisture on a constructed old-field ecosystem. The relative importance of main vs. interactive effects depended on the response variable and year. For example, total aboveground NPP was 60% greater in wet than in dry plots, but was unaffected by main or interactive effects of other variables. Soil nitrogen availability was also greater in wet than in dry plots, but was unaffected by [CO₂] or warming. In contrast, effects of [CO₂], warming, and soil moisture on leaf area index (LAI) in 2004, though significant, were strongly dependent on sampling date; in 2005, temporal variation in LAI was controlled by [CO₂] alone. Green tissue  [N] in 2004 was lower in elevated [CO₂] treatments than in ambient [CO₂], and was little affected by other variables. The importance of interactions between driving variables is dependent on the variable and year in question, and general patterns remain elusive.