The purpose of this study is to investigate the moderator role of judgements of learning (JOL), which is one of the metacognitive judgements, over the effects of being monolingual or bilingual on children's cognitive flexibility. A total of 75 bilingual and 75 monolingual children aged between 45 and 77 months committed to the study. Accuracy of JOL was evalauted by picture cards prepared by the researchers, cognitive flexibility was evaluated by DCCS tasks, and Turkish receptive language abilities were evaluated through TIFALDI-AD. A translated version of PPVT into Armenian was administered to bilinguals for evaluating their Armenian language abilities. A significant difference was found among equal bilinguals, non-equal bilinguals, and monolinguals in their cognitive flexibility levels and their accuracy of judgments of learning (JOL). Bilingual groups were found to perform higher in cognitive flexibility and accuracy of JOL when compared to monolingual groups. A multiple hierarchical regression was applied to investigate the moderator role of JOL in terms of what effects being bilingual or monolingual have on cognitive flexibility. Findings indicate that accuracy of JOL has a moderate effect in the relationship between cognitive flexibility and whether one is bilingual or monolingual. When the moderator role of the accuracy of JOL in the relationship between first stage scores of cognitive flexibility and being bilingual or monolingual was examined, it was observed that the participants of the bilingual group with a low accuracy of JOL had higher cognitive flexibility compared to the monolingual group, whereas there was no significant difference in cognitive flexibility level between the bilingual group with a higher accuracy of JOL and the monolingual group. When the moderator role of the accuracy of JOL in the relationship between the second stage scores of cognitive flexibility and being bilingual or monolingual was examined, it was observed that cognitive flexibility of the bilinguals was higher when compared to monolinguals in both groups with higher and lower accuracy of JOL. Findings were discussed relative to a representational redefinition explanation.