When collaboratively solving problems, children discuss information reliability, for example, whether claims are based on direct or indirect observation, termed as “metatalk”. Unlike English in which evidential marking is optional, languages with obligatory evidential marking such as Turkish, might provide children some advantages in communicating the reliability of their claims. The current preregistered online study investigated Turkish‐ and English‐speaking 3‐ and 5‐year‐old children's (N = 144) use of metatalk. The child and the experimenter (E) were asked to decide in which of the two houses a toy was hiding. One house had the toy's footprints. When E left the Zoom meeting, an informant told the child that the toy was in the other house without the footprints in three within‐subjects conditions. In the direct‐observation condition, the child witnessed the informant move the toy. In the indirect‐witness condition, the informant checked both houses and said that the toy was in the other house. In the indirect‐hearsay condition, the informant simply said that the toy was in the other house. When E returned, the child had to convince E about how they knew the toy was in the other house using metatalk (e.g., “I saw it move”). Turkish‐speaking children used metatalk more often than did English‐speaking children, especially in the direct‐observation condition. In the two indirect conditions, both groups of 5‐year‐olds were similar in their use of metatalk, but Turkish speaking 3‐year‐olds produced metatalk more often than did English‐speaking 3‐year‐olds. Thus, languages with obligatory evidential marking might facilitate children's collaborative reasoning.Research Highlights Children as young as 3 years of age can produce metatalk. Turkish‐speaking children produce metatalk more often than English‐speaking children. The difference between the two linguistic groups is more pronounced at age 3.