The traditional physical-based approach for estimating inflow design floods of large dams relies on the moisture maximization technique for estimating probable maximum precipitation (PMP). This procedure involves the direct product of maximum precipitable water and the highest precipitation efficiency derived from historical records. Nevertheless, this approach may overlook the influence of favorable atmospheric conditions that can transform maximum precipitable water into extreme precipitation amounts. Additionally, the formal moisture maximization approach prescribed by the World Meteorological Organization does not account for the long-term climatic trends of PMP components. This study investigates the relationship between seasonal maximum precipitation and corresponding precipitable water using gridded observations spanning the 1940–2021 period across Canada. Utilizing linear regression analysis, we observe a substantial proportion of grid cells showing non-significant relationship during summer, challenging the assumption of a linear conversion rate in moisture maximization-based PMP estimation procedures. Moreover, Mann–Kendall trend analysis reveals significant positive trends in seasonal maximum precipitable water, particularly pronounced during summer, indicating the potential inadequacy of the stationarity assumption for PMP estimation. We further investigate the suitability of additional atmospheric variables, including vertical integrated moisture divergence, vertical velocity, and convective available potential energy, alongside precipitable water. The analysis underscores the importance of considering regional variations in correlations for improving the accuracy of the moisture maximization procedure. These findings highlight the need for revisiting this commonly used procedure for PMP estimation.