Computer-based learning environments create the possibility of dynamic adaptation to address learner capabilities and user performance. Software algorithms, code-named Smart Tutor, for motor skill learning were developed and applied to an abstract environment for laparoscopic surgery (RapidFire). Smart Tutor dynamically adjusts the environment to minimize frustration and optimize learning conditions for all learners. This study compared the first generation RapidFire / Smart Tutor (RF / ST) to the Minimal Invasive Surgery Trainer Virtual Reality (MIST VR) system for laparoscopic performance improvement and level of frustration. Two groups of novice laparoscopic learners were assessed by pre- and post- training paper cutting exercise and subjective surveys. Users of both systems showed improvement of laparoscopic skills as measured by the paper cutting exercises. No differences were shown between groups for level of improvement. However, a significant difference was seen in the subjective ratings on the post-training survey with less frustration for the RF / ST training group. Important information was acquired for refinements of the Smart Tutor algorithms.