Reliable indices of abundance are crucial data for stock assessment. We applied a Vector-Autoregressive Spatio-Temporal Model (VAST) to conduct an index standardization using CPUE (catch-per-unit-effort) data of the striped marlin (Kajikia audax) in Western and Central North Pacific Ocean (WCNPO) during 1995 - 2020 collected from Taiwanese large-scale longline fishery logbook. CPUE standardization results between VAST and conventional generalized linear mixed models (GLMMs) were compared to evaluate the appropriate standardized CPUE index for the striped marlin stock assessment. VAST and conventional GLMMs were applied with year, season, vessel, spatial, and spatio-temporal effects as explanatory variables. Results indicated that VAST performs better than the GLMMs based on relatively higher R2 and lower Akaike Information Criterion (AIC) values. Relative abundance index derived from VAST indicated that the abundance index of striped marlin in WCNPO fluctuates overtime and has been increasing from 2018 to 2020. This study provides the basic information for the stock status of WCNPO striped marlin and could serve as a valuable input for future stock assessment if needed.