Shade tolerance, the minimum light requirement for plant survival, is a key trait for understanding community assembly and forest dynamics. However, it is poorly defined for tree species to date. Current methods of measuring shade tolerance vary considerably in their performance. For instance, some measures of shade tolerance are unreliable except under some specific conditions. Therefore, it is necessary to compare the performance of these methods to provide guidance of choosing appropriate shade tolerance measures in future studies. We collected a large dataset of light traits and other life history traits for 137 understory wood species in a subtropical forest and tested the performance of five commonly used shade-tolerance indices. Results showed that all the shade-tolerance measures, except the low-light abundance index, performed poorly in distinguishing and ranking shade tolerance of the tested species. The shade tolerance quantified by the low-light abundance was consistent with empirical classification of shade-tolerance/intolerance groups and successional seral stages of species. Comparison of the shade tolerance between trees of different diameter at breast height (DBH) or height classes further confirmed the reliability of low-light abundance. We conclude that low-light abundance is the most objective and practical of the five most commonly-used methods for measuring and ranking shade tolerance of understory wood species in our study forest, and likely in other forests as well. The simplicity of the method should greatly facilitate the assessment of light niche differentiation between species and thus contribute to understanding coexistence of tree species in forests.