OBJECTIVE: To investigate differences in the quality, confidence, and consistency of intraoperative surgical decision making (DM) and using functional neuroimaging expose decision systems that operators use. SUMMARY BACKGROUND DATA: Novices are hypothesized to use conscious analysis (effortful DM) leading to activation across the dorsolateral prefrontal cortex, whereas experts are expected to use unconscious automation (habitual DM) in which decisions are recognition-primed and prefrontal cortex independent. METHODS: A total of 22 subjects (10 medical student novices, 7 residents, and 5 attendings) reviewed simulated laparoscopic cholecystectomy videos, determined the next safest operative maneuver upon video termination (10 s), and reported decision confidence. Video paradigms either declared ("primed") or withheld ("unprimed") the next operative maneuver. Simultaneously, changes in cortical oxygenated hemoglobin and deoxygenated hemoglobin inferring prefrontal activation were recorded using Optical Topography. Decision confidence, consistency (primed vs unprimed), and quality (script concordance) were assessed. RESULTS: Attendings and residents were significantly more certain (P < 0.001), and decision quality was superior (script concordance: attendings = 90%, residents = 78.3%, and novices = 53.3%). Decision consistency was significantly superior in experts (P < 0.001) and residents (P < 0.05) than novices (P = 0.183). During unprimed DM, novices showed significant activation of the dorsolateral prefrontal cortex, whereas this activation pattern was not observed among residents and attendings. During primed DM, significant activation was not observed in any group. CONCLUSIONS: Expert DM is characterized by improved quality, consistency, and confidence. The findings imply attendings use a habitual decision system, whereas novices use an effortful approach under uncertainty. In the presence of operative cues (primes), novices disengage the prefrontal cortex and seem to accept the observed operative decision as correct.