This study contributes to the extant literature on the nexus among rice, maize and wheat production with agriculture gross domestic product (AGDP) of Pakistan. We use time series data from 1970 to 2017 and employ the Non-linear Autoregressive Distributed Lag (NARDL) model. Short run and long run shocks between the selected variables and result’s is checked through the co-integration and nonlinear error correction model.Autoregressive distributed lag bound testing approach for co-integration and to find the relationship between variables Granger causality test is applied.Our results confirm co-integration, positive shocks results show that rice, maize and wheat production have significantly influence on AGDP. The asymmetrically positive shocks of three crops have neutral effect on AGDP. While in symmetric results show the unidirectional effect between rice, maize production with AGDP and wheat production do not have ganger causality with AGDP. Finally, results depict that wheat, maize and rice production significantly contributes to agricultural GDP in the case of Pakistan.