There are two competing predictions for how experts will teach novices. According to one account, general rules and principles are better for equipping learners to perform the complex tasks that are characteristic of many domains of expertise. On the other hand, synthesizing a rule can be challenging for an expert, particularly in real-life environments with complex tasks, and learners may not be able to apply the rule to novel cases. In the present study, we examine a sample of six expert-novice pairs in the context of a robot-assisted surgical procedure in a teaching hospital. The nuances of this domain allow us to examine how an expert’s decision to teach inductively or deductively varies systematically across task complexity. We find initial evidence that experts in this domain primarily teach using specific case-by-case examples, as opposed to general rules and principles, regardless of the complexity of the surgical task.