Leaf Area Index(LAI)is a key biophysical variable for describing the canopy architecture.Currently,several global scale and long term LAI data sets have been produced.Validation of the LAI products is critical for their proper use in land surface models.However,existing studies were limited over flat terrain,the performance of the LAI data sets in mountainous area is still unknown.To fill this gap,the GEOV1,GLASS and MODIS LAI products were indirectly validated over mountainous area,southwestern China.The validation was conducted through the evaluation of the spatio-temporal continuity and the capability to capture phonological characteristics.The results indicate that LAI products shows worse spatio-temporal continuity in mountainous area than that in flat area.The best continuity is achieved by GLASS in our study areas and the worst is from GEOV1.GEOV1 and GLASS LAI product shows good spatial consistency,but MODIS is different from them.All products can hardly capture the details of the altitudinal distribution of biome types.The difference between the three products can be neglected over grassland but obvious over forests and cropland.GLASS LAI time series are smooth and continuous.GEOV1LAI time series have lots of gaps and MODIS LAI time series are very shaky.All three LAI products display unreasonable seasonality for evergreen needleleaf forests and crops.This study may facilitate the improvement of the quality of LAI product over mountainous areas.