Vehicle routing problems with time windows (VRPTW for short) is an NP-hard problem and an extension of vehicle routing problems. This study presented a heuristic algorithm (ABC-T) to deal with vehicle routing problems with time windows. The ABC-T algorithm employed the artificial bee colony algorithm (known as ABC) and improved the global search capacity with tournament selection strategy. In this algorithm, the employed bees generated the new solutions by the neighborhood search; then, based on the tournament selection strategy, the food source selection probability was settled. Finally, the new solution was generated by the onlooker bees. The tournament selection strategy was used here to enhance the global search capacity of the ABC. Compared with the state-of-the-art algorithms available in the literature, the effectiveness of this approach was showed on the Solomon’s R102 problem.