Ships avoid severe sea states for the safety of lives and property and to save energy. Therefore, the actual sea states encountered by ships differ from nature and are affected by human operational effects that we refer to as the “ship operational effect”. The evaluation of the ship operational effect is crucial for the rational design of structural strength of the hull. In this study, we derived the sea states encountered over 25 years in the North Atlantic, by considering storm avoidance by merchant ships. We then performed long-term predictions of the ship motions and wave loads of 75 merchant ships. Existing pioneering studies used automatic identification system (AIS)-based ship locations and timestamps to subsample wave hindcast data along ship tracks for a limited duration. Unlike existing studies, we used a statistical storm avoidance model of merchant ships to reconstruct the sea states, encountered over 25 years, in the North Atlantic. Owing to the ship operational effect, the long-term prediction values of roll motion, pitch motion, vertical bending moment amidships, and hydrodynamic pressure on bottom centerline amidships of merchant ships, at the probability level 10−8, were reduced by approximately 10–20%, and the coefficient of the ship operational effect was 0.8–0.9. Furthermore, we evaluated the uncertainties in the ship operational effect. The values deviated by as much as 6%, depending on the choice of wave hindcast products (TodaiWW3-NK, IOWAGA, and ERA5). The variation based on the parameter settings of the storm avoidance model (severity of the storm, radius of influence of the storm, and number of ship locations) was up to approximately 1 m at a probability level of 10−5 of the encountered sea states for 25 years.