In the western and central Pacific Ocean, upper strata waters exhibit highly dynamic oceanographic features under ENSO variability. This has been proved to be responsible for the dynamic change of both abundance and zonal distribution of skipjack tuna (Katsuwonus pelamis). Although causality has been suggested by researchers using physical–biological interaction models, cumulative evidence needs to be obtained and the tenability of assertion needs to be tested from an ecological habitat perspective, based on fisheries data. For purse seine fishery, the use of catch per unit effort (CPUE) as an indication of the abundance is confusing because of technical improvements over the whole exploitation history and unbalanced individual fishing characteristic of vessels. It is particularly interesting to discriminate between habitat characteristics in comparative scenarios of CPUE application. This study identified habitat traits based on a series of oceanographic factors from a global ocean reanalysis model. A comparison was conducted between two habitat models based on unprocessed purse seine CPUE and standardized CPUE considering fishing characteristics. The results suggest that standardized CPUE could model the regular zonal shift of habitat compatible with the observed fishing efforts transfer, and achieved better prediction capacity than unprocessed CPUE. Furthermore, the habitat of skipjack tuna was also characterized and linked with surface and subsurface thermal environment, ocean current, dissolved oxygen, biotic environment, and ENSO variability. The monthly-averaged habitat suitable index, derived from the optimal habitat model prediction, showed a significant linear relationship with the southern oscillation index, which suggested that El Ni?o episodes eventually provide more preferable habitat for skipjack tuna under ENSO variability.