The meteorological droughts are temporary, recurring events caused by insufficient precipitation. In this study, we considered two adjacent semi-arid and arid regions to estimate trends and meteorological droughts from the precipitation data. An artificial neural network (ANN) model is used to forecast precipitation at all stations. The geographical boundaries of these two regions are close to each other. Therefore, it is important to study an integrated trend analysis and meteorological drought forecasting. Modified Mann-Kendall's test and the standardized precipitation index (SPI) were used to estimate trend and drought at different time scales across all stations in the study area. March, May, June, summer, and yearly precipitation showed an increasing trend, while February showed a decreasing trend. The meteorological drought events and their severity significantly decrease with an increase in the time scale in observed and forecasted precipitation in both semi-arid and arid regions. For instance, SPI 3 showed the maximum droughts, while SPI 18 showed the minimum. The correlation coefficients and root mean square error (RMSE) between training, testing, and ANN modeled precipitation lies between 0.98 - 0.95, and 6.8 -18.4 mm across all stations in the semi-arid and arid regions, respectively. The minimum mean bias errors of less than 2 mm and 1 mm were estimated between the observed and validated precipitation data across all stations in the semi-arid and arid regions, respectively.