This paper proposes GAPS, a practical genetic algorithm based pathfinding system for conveniently analyzing various networks. To this end, the GAPS is developed through integration of the intuitive graphic user interface for network modeling, the database management system for managing the data generated in modeling and exploring procedures, and a simple genetic algorithm for analyzing a wide range of networks. Especially, previous genetic algorithms are not appropriate for analyzing the networks with many dead-ends where there are few feasible paths between the given two nodes, however, GAPS is based on the genetic algorithm with the fitness function appropriate for evaluating both feasible and infeasible paths, which enables GAPS to analyze a wide range of networks while maintaining the diversity of the population. The experiment results reveal that GAPS can be used to analyze both networks with many dead-ends and networks with few dead-ends conveniently, and GAPS has several advantages over the previous genetic algorithms for pathfinding problems.