The Good Behavior Game (GBG) is a classroom management intervention whereby students earn points toward a particular criterion to "win" the game. Typically, students know the criterion at the start of the game. However, this approach may cause behavior to deteriorate if, during the game, students believe that they have already won or lost. The current study investigated the effects of known and unknown criteria in a Welsh Year 4 (ages 8-9) classroom. Participants included three students who engaged in disruptive behavior (target students) and three who typically did not (nontarget peers). Tau-U and percentage nonoverlapping data (PND) effect size calculations indicated that both known and unknown criterion games were effective at reducing target students' disruptions to within the range of their nontarget peers, but neither game was more effective than the other. Teachers reported that they preferred playing the GBG with an unknown criterion; however, the students' preferences were mixed.