Evaluating evidence and restructuring beliefs based on anomalous evidence are fundamental aspects of scientific reasoning. These skills can be challenging for both children and adults, especially in domains where they possess inaccurate prior beliefs that can interfere with the acquisition of correct scientific information (e.g., heavier objects fall faster than light ones). Across two experiments, we examined the additive benefit of combining explanations with guided activities to promote conceptual change. In Experiment 1 (N = 238), 4- and 5-year-olds were randomly assigned to one of three conditions: guidance with explanations, guidance only, or baseline. The guided conditions varied only in the presence or absence of conceptual information (i.e., explanation about gravity). Pre- and posttest measures showed that children's predictions improved from both guided conditions compared to the baseline condition but did not significantly differ from each other. Experiment 2 (N = 80, 5-year-olds) included a delay test and assessed children's learning through the justification of their predictions. Although children's performance at the immediate posttest improved in both conditions, in the guidance only, children's performance returned to the pretest levels of understanding after the delay. Children in the guidance with explanations condition had greater understanding at posttest, retained this understanding long term, and transferred it to objects with the same weight. These findings highlight the role of explanations in aiding children's long-term learning from anomalous evidence in guided activities.