<p>This study investigates the impacts of climate variability, particularly temperature extremes and drought conditions, on wheat and rice production in Pakistan, a South Asian country facing significant agricultural challenges due to its diverse geography and climate. The primary objective of this research is to assess the relationship between climate variables (Land Surface Temperature (LST), Precipitation, and Standardized Precipitation Evapotranspiration Index (SPEI)) and crop yields for wheat and rice. The methodology integrates high-resolution climate data, including ERA5 LST and water availability datasets, with agricultural statistics from the Pakistan Bureau of Statistics. The analysis includes a time series evolution of climate variables and their correlation with crop yields, followed by the application of multiple regression models, machine learning techniques (including Polynomial Regression, Random Forest, and Support Vector Regression), and ridge regression for dimensionality reduction and model optimization. Results indicate that both wheat and rice yields exhibit an upward trend over the study period, with notable fluctuations attributed to rising temperatures and increasing drought frequency. The Random Forest model outperformed other methods, demonstrating high predictive accuracy for yield forecasting. The findings highlight the significant role of temperature in influencing crop productivity, while drought conditions, as indicated by SPEI, exert a negative impact on yield variability. This research contributes valuable insights into the climate-agriculture relationship in Pakistan and underscores the need for adaptive agricultural practices and climate-resilient strategies to mitigate the adverse effects of climate change on food security.</p>