Drought is a significant environmental challenge, particularly in semi-arid regions like the Bidar district, where understanding its temporal dynamics and predictive modeling was crucial for effective water resource control. This study used predictive models to forecast SPI values and examined changes in Standardized Precipitation Index (SPI) at various timescales across five stations in Bidar to assess drought patterns. To find trends, the Mann-Kendall test was used. While ARIMA and ANN models were applied for predictive analysis. Trend analysis revealed no statistically significant patterns across stations and timescales, although Basavakalyan and Bhalki exhibited marginally significant positive trends at longer timescales, suggesting a potential improvement in drought conditions that requires further scrutiny. In the predictive modeling phase, ANN models demonstrated strong performance during training by effectively capturing the nonlinearities in the data. However, ARIMA models outperformed ANN during testing, achieving higher accuracy and reliability for unseen datasets. These findings underscore the utility of ARIMA models for robust SPI prediction while highlighting the potential of ANN models with further optimization. ”By combining these methods, the study offers a comprehensive understanding of the drought dynamics in the region. It highlights the significance of integrating trend analysis with predictive modeling to enhance drought monitoring, planning and mitigation efforts in vulnerable areas such as Bidar district.”