<p>Sub-Himalayan West Bengal (SHWB) region experiences frequent pre-monsoon thunderstorms but lacks Doppler Weather Radar coverage, posing challenge for timely forecasting. To address this gap, this study aims to evaluate the performance of various thermodynamic indices and suggest suitable threshold values in order to enhance forecasting skill. The study is based on RS/RW data collected at 00 UTC (05:30 IST) daily during the storm field phase from March to May 2017–2018, which is exclusively available for the Jalpaiguri station (26.55°N, 88.71°E) in the entire SHWB. Optimal thresholds and relative forecast skills were determined through verification measures, like True Skill Statistic (TSS) and Heidke Skill Score for all the dichotomous predictors. We also combined TSS and Heidke scores into a Normalized Skill Score to establish a single optimal threshold. The thresholds were verified and validated using IMD station-level reports, ECMWF ERA5 CAPE climatology and INSAT-3D Cloud Brightness Temperature (CBT) images for five pre-monsoon thunderstorm cases. The results demonstrate that the BI, TTI and the Lowest 100&#xa0;hPa LI are the most effective for dichotomous forecasts. Other indices (DEW, HI, RH and SWEAT) perform well too, while prediction efficiency is notably low for indices KI, CAPE and DCI. The prescribed threshold values of thermodynamic indices in this study demonstrate reasonable effectiveness in predicting approaching thunderstorms, yet a single index cannot exclusively forecast them with lead time. Despite relatively low CAPE values (&lt; 1000&#xa0;J kg − 1) in the higher-elevation regions, thunderstorms can occur driven by dry convection and orographic lifting. The threshold values align with the CBT imagery-based nowcasting of intense convective activity during pre-monsoon months in the study area. The underpinning of this work could be enhanced with supplementary databases collected in subsequent years, improving the accuracy for future predictions.</p>