The relationship between children's reward-distribution judgments and their distribution strategies was investigated. Five-year-old children (N = 61) were presented with two stories where two characters made different numbers of origami stars. The children were asked to distribute different numbers of rewards to the characters: equal to (Middle-N), less than (Small-N), or more than (Large-N) the total number of stars in each story. Distribution strategies were categorized into two types: One-round, where rewards were distributed in one round only, and Cyclic, where the rewards were distributed in several cycles across the characters. In the Small-N of both stories (4 or 8 rewards), most children distributed rewards equally. When the number of rewards was 4, more than half used the One-round strategy, but when it was 8, more than half used the Cyclic strategy. In the Middle-N and Large-N conditions, most equal distributions used the Cyclic strategy, whereas almost all the proportional-equity distributions were associated with the One-round strategy, and most ordinal-equity distributions used the Cyclic strategy. The relationships between automatic/controlled reward-distribution judgments and distribution strategies were discussed.