Variation in spring precipitation in Pakistan (SPP) exerts tremendous socio-economic impacts, particularly by sustaining local agricultural practices and ensuring food security in this arid to semi-arid region. Accurate seasonal prediction of the SPP is essential for alleviating socio-economic susceptibilities. However, understanding the physical cause and improving the predictability of interannual SPP variations remain critical yet challenging tasks. Here, we identify preceding winter sea surface temperature (SST) in the extratropical and tropical North Pacific as robust and independent seasonal predictors of the SPP, which are corresponding to the North Pacific Victoria Mode (VM) and El Ni & ntilde;o-Southern Oscillation (ENSO), respectively. These SST predictors significantly influence the SPP pattern by inducing atmospheric and oceanic teleconnection Rossby wave train, respectively. The physical mechanisms based on observations are supported by simulations from the Atmospheric Model Intercomparison Project (AMIP) within the Coupled Model Intercomparison Project Phase 6 (CMIP6). By incorporating these SST anomalies as key physical drivers of interannual SPP variability, we develop a physics-based statistical model for the seasonal prediction of SPP. The physics-informed prediction model based on the SST predictors is capable of forecasting the SPP at one-season lead, which exhibits a high cross-validated hindcast skill at 0.76 during 1979-2020. These findings enhance seasonal forecasting skill for the SPP and offer substantial benefits to the agriculture-dependent communities in the region.