The objectives of this work were to clarify the spatial distribution pattern for the Fusarium wilt disease in the banana field and the amounts of Fusarium oxyporum f. sp. cubense population inside the banana plants. An investigation was conducted on the diseased plants in banana field. The spatial distribution pattern of the Fusarium wilt disease in banana field was studied by determination of frequency distribution, aggregated indices, the Taylor's power law and Iwao's regression equation. The reasons for the aggregation of the Fusarium wilt disease in banana field was analyzed according to estimation by aggregation mean λ. Meanwhile, comparative investigation was conducted on the amounts of the pathogen inside the healthy plants and diseased plants. Results indicated that the diseased plants in banana field belong to negative-binomial distribution and neyman distribution, and the aggregated indices of diseased plants met C>1, I>0, m^*/m>1, CA>0, K>0, so they presented pattern of aggregation distribution, and the fundamental component was the individual that was mutually attractive, and the individual colonies characterized with uniform distribution. The degree of aggregation increased with the population density of diseased plants. The rate of relative aggregation change of fundamental component was (11.0962+0.1752)m, which changed with the population density of Fusarium wilt disease in banana field. The greater the population density of diseased plants, the greater the change rate of relative aggregation degree. Aggregation mean λ value indicated that the aggregation might be due to the combined actions of the environmental factors. The optimal sampling formula was built up. The sample number (n) negatively decreased with the pathogen indeces increasing, and it could be estimated with the parameters of t=1.96, deviation D=0.1-0.4. The sequency sampling procedure was based on the equation. It was necessary to take actions to control Fusarium wilt disease when the plant numbers were higher than the upper limit numbers according the index list for sequential sampling of infected plants in banana field. Chessboard type and diagonal type sampling were the best method for sampling technique. The application of spatial pattern to data transfer was discussed. In addition, the amounts of the pathogen in the bulbs were more than those in roots and leaves. The pathogens in different parts of banana plants were significantly different due to the disease severity of the fields and sampling parts of banana plants.