This work analyzes the spatial distribution of the Bactrocera carambolae occurrence between 2014 and 2019 in Amapá, Brazil. They monitored the quarantine pest present using traps distributed throughout the state. The technicians checked traps every fifteen days, recorded the occurrence, and sent the information to MAPA. To analyze the spatial distribution, we evaluated the degree of randomness of the traps using the non-homogeneous K function. Subsequently, we applied the quartic kernel function on the occurrences and the simulated occurrences, considering the non-homogeneous phenomenon and having the population per census tracts and highways as covariates. The results showed that we could use the method to monitor the pest on a regional scale.