Nocturnal sap flow (Qn) is an integral component of plant physiological processes and ecosystem water cycling. However, global patterns in the relative contribution of nocturnal to total daily sap flux (Qn%) remain poorly understood, limiting our ability to predict plant and ecosystem water use under climate change. To address this gap, we leveraged a global sap flow network (SAPFLUXNET) to assemble a dataset of Qn% comprising 1,366 woody individuals from 79 species across 121 sites worldwide. We used this dataset to examine Qn% variations and associated drivers at the global scale. Across geographic gradients, Qn% exhibited a hump-shaped relationship with latitude, averaging 12.6% across all woody plants. Gymnosperms exhibited significantly higher Qn% values than angiosperms, highlighting the role of plant functional strategies. Qn% also varied among biomes-being highest in desert regions and lowest in tropical and boreal forests. Partitioning nocturnal sap flow using the forecasted refilling method revealed that Qnwas globally dominated by stem water refilling, which accounted for 82% of Qn. Under arid conditions, however, the relative contribution of nocturnal transpiration to Qnincreased, and the sensitivity of Qn% to short-term water stress was amplified. Further, using a linear mixed effects model, we identified two previously under-appreciated factors-species identity and nighttime length-as crucial drivers of the Qn% variation at the global scale. Together, these findings enhance our understanding of nocturnal sap fluxes and offer key insights for modeling spatiotemporal patterns of plant water relations and ecohydrological dynamics in a changing environment.