Accurate estimation of standing tree volume is vital for effective harvest planning, revenue projection, and ensuring ecological sustainability. However, the commonly used allometric equations in Nepal often overlook species-specific characteristics, resulting in significant discrepancies between estimated and actual volumes. This study compares values derived from standard allometric equations with actual measurements taken after felling in Sal (Shorea robusta) forests of Kapilvastu, Nepal. A total of 1588 trees were analysed to assess the extent and causes of volume estimation errors. The results reveal significant differences between estimated and actual tree heights and volumes, with estimated volumes often exceeding the actual timber yields. These discrepancies are particularly observed in Class I trees, suggesting limitations in current classbased volume adjustment factors. The study also identifies key variables contributing to these inaccuracies, including tree size, measurement method, and operator variability. Statistical analyses, including paired t-tests and ANOVA, confirm the systematic nature of these differences. The findings highlight the need to update or localise allometric models to improve accuracy in volume prediction. This research provides empirical evidence for refining volume estimation practices and supports better-informed decision-making in forest harvesting, ultimately contributing to more sustainable forest management strategies in Nepal and similar ecological settings.