Two experiments examined the role of expertise, consensus, and informational valence on children's acceptance of informant testimony about the quality of work produced by a target child. In Experiment 1, 96 4- to 5.9-year-olds and 6- to 8-year-olds were told about an expert who gave a positive or negative assessment of art or music that was contradicted by one layperson or a consensus of three laypersons. Generally, participants endorsed positive assessments as correct irrespective of expertise and consensus, but older children were more likely than younger children to want to learn from the expert in the future. To examine whether reluctance to accept expertise was due to the negative quality of the information, the expert in Experiment 2 simply stated that additional work was needed. Both age groups selected the expert as correct and reported wanting to learn from the expert in the future. Contributions to social learning models are discussed.