Stem CO 2 efflux (E s ) has been estimated from a temperature-related equation, but sap flux often affects measurements of E s , which leads to misunderstanding real stem respiration. In order to observe the relationship between E s and stem temperature and to analyze the effect of sap velocity on E s , stem temperature, E s and sap flux were measured from a subtropical Schima superba plantation in South China on three trees for consecutive 3 days in July and October 2009. Stem temperature, E s and sap velocity were significantly higher in July than in October. Stem temperature could explain 17-41 and 54-75% variations of E s in July and October, respectively. A negative relationship between E s and stem temperature was found during 1800-2300 hours in July. The daytime E s was 9.2, 4.3 and 2.4% higher than the predicted for three trees in July, and this occurred only on Tree 1 in October. Sap velocity was positively correlated with E s for three trees in July, and the increase of E s with the increase of sap velocity was only observed on Tree 1 in October. These results demonstrated that the occurrence of sap flux could account for the increase of daytime E s , and the effect of sap velocity on E s varied with the seasons from the S. superba stem.