1. Individual effects of co‐occurring global change factors on net primary productivity (NPP) have been widely studied; however, their interactive effects remain highly debated.

2. Here, we conducted a global meta‐analysis based on 919 multifactor observations from 120 published studies to examine the interactive effects on NPP of global change factors including elevated [CO₂], warming, nitrogen addition, irrigation, drought, and changes in species diversity.

3. On average, of the factors studied, six pairs of factors had additive and two pairs had synergistic interactions. Importantly, some of those interaction types changed over time and with treatment intensity. The synergistic interaction between elevated [CO₂] and nitrogen addition became additive at high nitrogen addition rates, whereas the synergistic interaction between irrigation and warming diminished at higher temperatures. Over time, the additive effect between elevated [CO₂] and increased species richness switched to synergistic. Other global change factor pairs—including elevated [CO₂] and warming, nitrogen addition and increased richness, irrigation and N addition, as well as drought and increased richness—remained additive regardless of their treatment intensity or experimental duration. Interaction types of those global change factor pairs did not vary with ecosystem types assessed in our study.

4. Synthesis. Our results suggest that assumptions of static effects through time or ignoring treatment intensity effects will provide inaccurate predictions of the interactive effects of global change factors on terrestrial NPP. Understanding the context‐dependent nature of interactive effects is crucial for validating Earth system models and predicting future NPP responses to co‐occurring global change drivers.

We searched peer-reviewed journal articles that evaluated the response of plant growth or ecosystem NPP under experimental manipulations of multiple global change factors using the Web of Science and Google Scholar up to 1^st August 2018. Different keyword combinations were used for the search, such as “(climate change OR elevated [CO₂] OR warming OR temperature OR nitrogen addition OR fertilization OR irrigation OR precipitation manipulation OR drought OR diversity OR richness OR mixture OR interaction effect) AND (plant growth OR production OR productivity OR biomass production)”. Only primary studies that satisfied the following criteria were included in this meta-analysis: 1) at least a 2 × 2 full-factorial design was used to examine the effects of global change factors including elevated [CO₂], warming, N deposition, irrigation, and/or drought, and change in species diversity; 2) at least one component of production (e.g., aboveground, belowground, or total) was examined under all treatments and the control at the same temporal and spatial scales; 3) the plots used for all treatments were of the same ecosystem type and environmental condition as the control; and 4) the mean and sample size of the selected variables were available or could be calculated from data provided in the publications. All original data were extracted from the text, tables, figures, and appendices of the publications. When data were graphically presented, SigmaScanPro version 5 (Systat Software Inc., Point Richmond, CA, USA) was used to obtain numeric data. Furthermore, data regarding experimental duration (years), treatment intensity (i.e., increased or decreased quantities applied in the treatment as compared with the control), ecosystem type (forest, grassland, cropland, tundra, wetland), production component (aboveground, belowground, or total), and growth form (woody or herbaceous), were also recorded from detailed descriptions in the Materials and Methods sections of papers included in the analysis.