Four monitoring wells in the Metro Atlanta region were used in this study to investigate how climate influences groundwater variability in unconfined aquifers in the Piedmont Physiographic Province. Antecedent groundwater levels and short- and long-term antecedent climatic variables (temperature and precipitation) were used as predictors in the development of multiple linear regression models. Variable selection was governed by the Bayesian Information Criterion (BIC), which balanced prediction accuracy and model complexity. Strong model performance is demonstrated by the results, with R2 values exceeding 0.9 for all models which show that both short-term rainfall events and long-term climatic patterns play a crucial role in predicting groundwater levels. The results highlight the importance of antecedent groundwater levels and climatic factors in capturing groundwater dynamics under various site settings.