Research demonstrates the importance of social interaction for children. Two important aspects of social interaction are conversation and scaffolding. Previous work has shown that children benefit from conversation with parents. However, no research has examined if children learn from conversation equally when interacting with different partners, so it is unclear how that learning may be different or similar to how children learn from parents. Even though it is clear that conversation is important for how children learn, no one has looked at conversation using the framework of scaffolding. Scaffolding is an ongoing interaction between a child and some other speaker where the other speaker provides support contingent on the child’s progress (Bruner, 1975). Because conversation is an essential part of scaffolding, it is important to understand what potential differences exist between different conversation partners to illuminate differences that may exist in scaffolding. Previous scaffolding research has found that peers scaffold differently than parents; in most cases, they are not as effective as mothers and they exhibit fewer scaffolding-like behaviors. Much less work examines how siblings scaffold; siblings are similar to peers but may be more familiar with the target child. However, there are several methodological problems with these studies, such as the mother being present or comparing children in different settings using different stimulus items. Further, no research has examined what is responsible for the fact that some people can scaffold better than others. In order to understand how children learn differently from different people, we need to know what is responsible for the difference in scaffolding. This research explores whether children can benefit equally from conversation with people other than their parents and seeks to explain why differences in scaffolding exist between partners. Study participants included ethnically diverse 3- and 4-year-old children recruited in local Merced County preschools. In Study 1, 144 participants engaged in a free play interaction with either their parent, older sibling, or other adult. Examining their conversations, results indicate that children seem to vary in how they benefit from conversation depending on their partner. In particular, children asked more questions of parents and other adults than siblings, but siblings provided more informative answers than parents and equally to other adults. Further, parents and other adults turned the question back to children more often than siblings did. In Study 2, the erroneous utterances and how partners responded to these were investigated in the same 144 children who participated in a free play task with either their parent, older sibling, or other adult. Results indicate that parents and other adults provided more reformulations than siblings did. This confirms that children do not equally benefit from conversation with different partners. In both Study 1 and Study 2, sibling age spacing does not seem to affect their interaction with children during conversation. In Study 3, 92 participants engaged in a building task with either their parent, older sibling, or older peer (age-matched to siblings). Scaffolding and orienting behaviors, as well as appropriate shifts in instruction, were investigated. Parents used more scaffolding and orienting behaviors than both siblings and peers. They also made more appropriate shifts during the course of interaction. Siblings and peers did not differ in the amount of scaffolding and orienting behaviors. They also did not differ in the amount of appropriate shifts made. In addition, children performed equally at independent post test regardless of who their teaching partner was, with 4-year-olds outperforming 3-year-olds. Because Study 2 confounded motivation with familiarity within siblings and parents, Study 3 introduced a control for motivation. In Study 3, 63 participants engaged in a building task with either their parent or older sibling. However, unlike Study 2, siblings were told that they would receive a reward if the child performed well during post test. Scaffolding and orienting behaviors, as well as appropriate shifts in instruction, were investigated. Despite introducing a reward, no effects were found between siblings’ scaffolding and orienting behaviors in this study and Study 3. Confirming the findings from Study 2, parents used more scaffolding and orienting behaviors, and made more appropriate shifts than siblings did. No effect of teaching partner was found with respect to children’s independent post test performance. Consistent with previous studies, sibling age spacing did not have an effect on scaffolding behaviors in both Study 2 and Study 3. Overall, the results of this research suggest that children learn differently from different people, and the benefits of parent-child interaction may not be as beneficial with other types of partners. Knowledgeability, familiarity, and motivation all seem to play a role in why people scaffold differently. This research adds importantly to both the conversation and scaffolding literatures, and these findings have implications for current theories of how children learn, parenting practices, and classroom application.