This paper proposes a fuzzy controller design using group-crossover particle swarm optimization (GCPSO) algorithm. The GCPSO uses a group-based framework for defining particle neighborhood topology and incorporating crossover operation into particle swarm optimization. The GCPSO dynamically forms different groups for selecting parents in crossover operations, particle updates and replacements. The objective of GCPSO is to improve fuzzy control accuracy. Comparisons with different population-based optimizations on truck reversing control problem demonstrate the performance of GCPSO algorithm.