Lifeboats are essential life-saving equipment for many commercial (and military) vessels and offshore platforms. Their coxswains usually learn to maneuver on calm waters at a nautical school for safety reasons. Simulator training could avoid the environmental constraints that limit training with an actual lifeboat (e.g., frozen water) while enlarging the experience (e.g., high winds). However, the agencies that regulate training for lifeboats have little available evidence of simulator effectiveness since little research has been conducted in this domain (Power-Macdonald, MacKinnon, Simões Ré, Power & Baker, 2011). As a first step in an ongoing investigation of simulator training for emergency operations of a lifeboat, we conducted a behavioral experiment to assess transfer of training of maneuvers practiced in a simulator to control of a lifeboat in benign environmental conditions. To reduce the effects that natural abilities and prior experience could have on the experimental outcome, all participants initially attempted a basic maneuvering course with the lifeboat. They were then paired by their performance. One member of each pair, chosen at random, subsequently received simulator training. After all members of the simulator trained group were trained to criterion on a virtual representation of the test course, the members of both groups individually completed the test course to criterion on the water with the actual lifeboat. The test course involved a sequence of challenging maneuvers. Ten participants in each group completed the test course. In this paper, we report learning in the simulator, and we compare initial transfer of training and improvement with practice to criterion by each treatment group. A transfer effectiveness ratio (TER) is calculated. In addition, we report the effect of practicing to criterion within the simulator on training transfer. We discuss the implications of the results for lifeboat training regulators, as well as their relevance to the broader training community.