Using dendroecological methods and empirical biomass equations, we calculated the biomass and net primary productivity (NPP) of three coniferous forests (Pinus armandii natural forest, Larix principis-rupprechtii plantation, Pinus tablaeformis plantation) in Liupan Mountains of Ningxia, northwest China. Then we analyzed their relation with climatic factors, including monthly air temperature and precipitation. The results indicated that the interannual variation of biomass (t/hm²) of the three forests studied follows the logistic equation, with the order of L. principis-rupprechtii plantation (141.96) > P. armandii natural forest (130.99) > P. tablaeformis plantation (123.29). The interannual variation of NPP of P. armandii natural forest showed a trend of "fast increase-slow increase-slow decrease"; while the variation of NPP of other two plantations showed a trend of "rapid rising" with some fluctuations. There is a remarkable difference among the NPP (t·hm^-2·a^-1) of the three forests, with the order of L. principis-rupprechtii plantation (6.72)>P.tablaeformis plantation (5.76) > P. armandii natural forest (2.66). With increasing annual precipitation, the NPP varied much differently among the three forests, i.e., the NPP of P. armandii forest increased very slowly, but the NPP of other two plantations firstly increased and then decreased. However, in the driest or the rainfall-richest years, the NPPs of all three forests tended to their same lower values, but probably with different stresses of drought or low temperatur. The NPP response to climatic factors presnted a time lag and acted differently among the three forests. The NPP of P.armandii forest was significantly and negatively correlated with the precipitation in November of previous year and that in November and September of current year; while the NPP of P. tablaeformis plantation was significantly correlated with the precipitation in September of previous year and that in April in current year, and the NPP of L.principis-rupprechtii plantation was significantly and positively correlated with the precipitation in September of both previous and current year. The NPP of all three forests was significantly correlated with the mean temperature and minimum temperature in June and March of current year, and the temperature in June of previous year. However, the NPP response to temperature differently among the three forests, the NPP of P.armandii forest was more or less negatively correlated with the temperature in each month of growing season; while the NPP of P.tablaeformis and L.principis-rupprechtii forests was more or less positively correlated with the temperature in each month of gowth season.