We analyzed the spatial distribution of tuna long-line fishing grounds in Atlantic using geostatistics. The data consisted of Japanese tuna long-line fishing logs from the Atlantic Ocean between 1982–2008. There was a strong spatial correlation for the big-eye tuna long-line fishing ground. The correlation was significant in June, October, and November(<0.01). The exponential model was best able to explain the spatial variability across the fishing ground (related coefficients between 0.6 and 0.9). Based on the model forecast, there were two patterns across the fishing ground: One cyclical changed between the mid-Atlantic and America, and the Second cyclical changed between the mid-Atlantic and Africa. According to the forecast results, geostatistics are better for fishing ground distribution forecasting. However, the predicted CPUE was higher than the measured value, likely because of differences in resource status. Our results can be used as a guideline for the development of the big-eye tuna fishery and exploration of long-line fishing grounds.