Purpose The aims of ideal preoperative informed consent include educating the patient adequately to enable an autonomous decision to be made without causing undue anxiety. We study how the paternalistic and nonpaternalistic approaches meet this ideal. The influence of the new patient consent forms is also assessed. Methods Two cycles of a prospective clinical audit are presented. An assessment of relevant patient knowledge was performed by patient interview. Visual analogue scales were used to quantify patient anxiety. Results The first cycle, examining a paternalistic approach, demonstrated: 37% of patients understood what a cataract was and 48% understood what surgery involved. 48% misunderstood that cataract surgery was completely risk free. In total, 80% of patients undergoing second eye surgery believed that it was completely risk-free. Average anxiety visual analogue scores (VAS) for cataract surgery were low (2.89). The second cycle, examining the nonpaternalistic approach combined with the implementation of new consent forms showed that, despite more explicit repeated preoperative consent: 39% of patients understood correctly what a cataract was, 28% understood what surgery involved and 43% misunderstood that surgery was completely risk-free. All patients undergoing second eye surgery thought that it was riskfree. The average anxiety VAS for cataract surgery were moderate (5.00). Conclusions Both paternalistic and nonpaternalistic approaches to informed consent are inadequate in meeting the demands of the ideal informed consent. The new patient consent forms appear to have little effect in influencing patient knowledge about their surgery. Patients undergoing second eye surgery often have an overoptimistic view of cataract surgery.