How the rapidly developing livestock grazing affects grassland aboveground net primary productivity (ANPP) is uncertain. The grazing optimization hypothesis (GOH) predicts that appropriate grazing pressure benefits ANPP and that the response of ANPP to grazing intensity follows a single-peaked pattern, but it remains controversial. Here, we conducted a global meta-analysis with 376 pairs of observations for ANPP to investigate the responses of grassland ANPP to stocking rate and forage consumption rate (FCR). Results showed that there was no significant relationship between stocking rate and ANPP response, primarily because the effect of stocking rate varied according to climatic conditions. FCRs were calculated using three methods (FCR I, II, and III), with FCR III providing a precise estimate of the actual FCR. ANPP responded unimodally to FCR III on the global scale, with the ANPP response peaking (12 % increase) at the optimal grazing intensity (FCR III = 58 %). The results for FCR I and FCR II also corroborated the single-peaked effect of the actual FCR. Based on our preliminary evidence, we propose a hypothesis that the single-peaked effect of FCR is consistent on a global scale, but this needs to be confirmed by further research. The unimodal relationship between ANPP responses and grazing intensity we found supports the GOH and informs win-win outcomes for grassland productivity and livestock production.