Nutrient resorption in plants influences nutrient availability and cycling and is a key process in biogeochemical models. Improved estimates of resorption parameters are needed for predicting long-term primary productivity and for improving such models. Currently, most models assume a value of 50% resorption for nitrogen (N) and phosphorus (P) and lack resorption data for other nutrients and for specific vegetation types. We provide global estimates of resorption efficiencies and nutrient concentrations for carbon (C), N, and P and the first global-scale estimates for essential nutrients such as potassium (K), calcium (Ca), and magnesium (Mg). We also examine leaf mass loss during senescence (LML) globally and for each plant type, thus defining a mass loss correction factor (MLCF) needed to quantify unbiased resorption values. We used a global meta-analysis of 86 studies and ~1,000 data points across climates for green and senesced leaves in six plant types: ferns, forbs, graminoids, conifers, and evergreen and deciduous woody angiosperms. In general, N and P resorption differed significantly from the commonly used global value of 50% (62.1 and 64.9%, respectively; P<0.05). Ca, C, and Mg showed lower average resorptions of 10.9, 23.2, and 28.6%, respectively, while K had the highest resorption, at 70.1%. We also found that resorption of all nutrients except Ca depended on leaf nutrient-status; globally, C, N, P, K, and Mg showed a decrease in resorption with increased nutrient status. On average, global leaf mass loss was 24.2%. Overall, our resorption data differ substantially from commonly assumed values and should help improve ecological theory and biogeochemical and land-surface models.