Abstract: The main goal of zone based school enrollment method is reducing the student’s home to school distance. So, among its variants, student will be allocated to the school that its location is near the student’s home location. Meanwhile, one biggest problem in this method is the number of out of zone students which means of the student that is accepted in school that its location is outside his own zone. This situation occurs, especially when the zoning is determined statically. In this research, we propose dynamic zone based school enrollment model to meet the variations in school location distribution, student location distribution and school capacity. This model is developed based on k-means clustering method with several modifications. This proposed model then is implemented into zone based school enrollment simulation to measure its performance. In this simulation, the performance of this proposed model is also compared with the existing static one. In this simulation, adjusted parameters are school location distribution, student location distribution, school capacity, number of schools and number of students. Meanwhile, the observed parameters are average student’s home to school distance and the success ratio. Based on the test result, there are several findings. The proposed model produces higher success ratio rather than the existing static method does. In the other side, the existing static method produces lower average student’s home to school distance rather than the proposed model does. The number of schools and the average school capacity has positive correlation with the success ratio. The number of students has negative correlation with the success ratio. These adjusted variables do not affect the average student’s home to school distance.