Using the Yale Interactive Biosphere (YIBs) model, as well as the data from the FLUXNET sites with measurement records longer than 8 years, this study examines the impacts of multiple meteorological variables on the interannual variability of gross primary productivity (GPP), and identifies the dominant drivers of GPP variability for different plant functional types. The results showed that the change in photosynthetically active radiation is the main driving factor of GPP interannual variability for deciduous broadleaf forests and evergreen needleleaf forests, contributing 80% of the variability for these forest types. The change of relative humidity is the main driving factor of the GPP interannual variability for cropland, contributing 65% of its GPP variability. Temperature is an important factor for shrubland, though its contribution rate is moderate. The GPP interannual variability of grassland shows large uncertainties among sites without uniform driving factors. Our results show that meteorological variables are the dominant drivers for the interannual variability of global terrestrial GPP. With the future climate change, more frequent climate extremes may further increase the GPP interannual variability.