Statistical sampling theory provides a robust framework for conducting VVPAT-based audits of electronic voting machines (EVMs) to ensure the accuracy and reliability of election results. This article analyzes the current VVPAT audit protocol prescribed by the Supreme Court of India and highlights its shortcomings. It emphasizes the need for a statistically sound approach that defines the target population, specifies clear acceptance criteria, and ensures that the sample size is adequate to detect election-altering malfunctions or manipulations with high probability. The article provides concrete recommendations for improving the VVPAT audit process and preserving the integrity of the electoral system.