<p>Abstract</p> <p>Background</p> <p>Simulation is increasingly used for teaching medical procedures. The goal of this study was to assess learner preferences for how simulators should be used in a procedural curriculum.</p> <p>Methods</p> <p>A 26-item survey was constructed to assess the optimal use of simulators for the teaching of medical procedures in an internal medicine residency curriculum. Survey domains were generated independently by two investigators and validated by an expert panel (n = 7). Final survey items were revised based on pilot survey and distributed to 128 internal medicine residents.</p> <p>Results</p> <p>Of the 128 residents surveyed, 106 (83%) responded. Most responders felt that simulators should be used to learn technical skills (94%), refine technical skills (84%), and acquire procedural teaching skills (87%).</p> <p>Respondents felt that procedures most effectively taught by simulators include: central venous catheterization, thoracentesis, intubation, lumbar puncture, and paracentesis. The majority of learners felt that teaching should be done early in residency (97%).</p> <p>With regards to course format, 62% of respondents felt that no more than 3-4 learners per simulator and an instructor to learner ratio of 1:3-4 would be acceptable.</p> <p>The majority felt that the role of instructors should include demonstration of technique (92%), observe learner techniques (92%), teach evidence behind procedural steps (84%) and provide feedback (89%). Commonly cited barriers to procedural teaching were limitations in time, number of instructors and simulators, and lack of realism of some simulators.</p> <p>Conclusions</p> <p>Our results suggest that residents value simulator-based procedural teaching in the form of small-group sessions. Simulators should be an integral part of medical procedural education.</p>