Abstract: The seasonal changes in LAI were respectively measured using digital hemispherical photography (DHP), LAI-2000 plant canopy analyzer (indirect methods) and litter collection method (direct method) in spruce-fir valley forest in the Xiaoxing’an Mountains, China. In addition, we constructed the relationships between LAI from direct method and that from indirect methods in different seasons. The results showed that LAI from DHP underestimated LAI from direct method by 40%-48% in the whole study period, and the underestimation range for LAI-2000 plant canopy analyzer was 15%-26%. LAI from direct method and that from DHP or LAI-2000 plant canopy analyzer all correlated significantly in different seasons, and the relationships between LAI from direct method and that from DHP or LAI-2000 plant canopy analyzer all could be merged into three classified prediction models after testing, i.e., A, B and C, which were able to predict LAI in May and November, June, September and October, and July and August, respectively. This study laid a foundation for effective and accurate measurement of seasonal changes in LAI for coniferous forest in subsequent studies.