Previous studies have demonstrated that the majority of cancer patients fail to achieve their desired level of participation during treatment decision making. However, it is unclear whether this failure affects patient well-being and satisfaction. Furthermore, whilst shared decision making is currently espoused as the preferred model for doctor patient relations, little empirical evidence exists showing it has beneficial effects for patients. We aimed to evaluate the impact of shared decision making and the achievement of preferred role on patient anxiety, recall of information, and satisfaction. Patients with cancer indicated their preferred level of participation in decision making and preferences for information and emotional support prior to their initial consultation with an oncologist. Anxiety was assessed prior to and immediately after the consultation and recall seven days after the consultation. Anxiety was reassessed at two-weeks post-consultation at which time patients also reported their satisfaction with the consultation and perceived role of participation in treatment decision making. Satisfaction with the information and emotional support received was also evaluated. Of the 233 patients available for analysis, a match between preferred and perceived roles was found for over one-third of patients (34%), with 29% more active and 37% less involved in decision making than preferred. Multivariate analyses demonstrated that role mismatch significantly predicted changes in anxiety levels from pre to immediate postconsultation (p = 0.03). However, irrespective of preferred role in decision making, perceived role, but not role mismatch, significantly and independently predicted satisfaction with both the consultation (p = 0.0005) and the amount of information and emotional support received from the doctor (p = 0.004). Patients who reported a shared role in decision making were most satisfied with the consultation and with the information about treatment and emotional support received. Those who reported that either themselves or the doctor exclusively made the decision were least satisfied. These findings underscore the pre-eminence of the shared decision making model and suggest that encouraging participation may be the safest standard approach. Doctor - as well as patient - based interventions are required to promote patient participation.