The Particle-In-Cell (PIC) method is effectively used in many scientific simulation codes. In order to optimize the performance of the PIC approach, data locality is required. This relies on efficient sorting algorithms. We present a bucket sort algorithm with small memory footprint for the PIC method targeting Graphics Processing Units (GPUs). Our sorting algorithm shows an increased performance with the amount of storage provided and with the orderliness of the particles. For our application where particles are presorted it performs better and requires less memory than other sorting algorithms in the literature. The overall PIC algorithm performs at its best if the sorting is applied.