Seventy-one patients were randomly allocated to undergo either computer-navigated or conventional arthroplasty. A statistically significant improvement in alignment was seen in the computer-navigated cohort. Five-year functional outcome was assessed using the Knee Society, Short Form-36, Western Ontario and McMaster Universities Osteoarthritis Index, and a patient satisfaction score. At 5 years, 46 patients were available for assessment (24 navigated and 22 conventional knees). No patients had undergone revision. No statistically significant difference was seen in any component of any measure of outcome between navigated and conventional cohorts. Longitudinal data showed function to be well maintained with no difference in functional score between 2 and 5 years in either cohort. Despite achieving better alignment, 5 years postoperatively, the functional outcome with computer-navigated knee arthroplasty appears to be no different to that implanted using a conventional jig-based technique.