Annual catch rates and catches are important fishery data to assess fish population dynamics. However, these data of sharks have an issue of under-reporting. To solve the issue, we standardized nominal CPUEs of shortfin mako caught by Japanese longline fisheries in the Indian Ocean from 1993 to 2018 using three observation error models (zero-inflated Poisson model: ZIP, negative binomial model: NB, and Poisson model: PO) with logbook data after filtering the data. The NB with full explanatory variables was selected by AIC as the most parsimonious model. The estimated annual catch rates (standardized CPUE) showed a decreasing trends with large fluctuations from the beginning of 1990s until 2009, and then they showed a slight increase trends. We also estimated the annual catches from 1993 to 2018 using the estimated annual catch rates and total fishing effort of Japanese longline fisheries operated in the Indian Ocean. The estimated catches increased in the beginning of 1990s and reached at peak in 1996 due to high fishing effort and catch rates. The estimated catches therefore gradually decreased due to the decrease in the fishing effort and lower catch rates. These annual catch rates and catches would be useful for the next stock assessment of shortfin mako in the Indian Ocean.