334 patients scheduled for a total knee arthroplasty were randomized to receive 6 weeks of therapist-led physiotherapy or home-based physiotherapy. The primary outcome of interest was Oxford Knee Score (OKS) at 1 year post-operation. Secondary outcomes of interest included the worst and average pain scores, OKS scores, EQ-5D scores, ability to perform tasks and patients satisfaction. Follow up was performed at 6 weeks, 6 months and 1 year post-operation. OKS throughout 1 year follow-up was significantly in favour of the therapist-led group. Patient satisfaction with pain relief, the ability to perform daily functional tasks and the ability to perform heavy functional tasks were significantly in favour of the therapist-led group. No differences in worst or average pain score or EQ-5D scores were observed between groups. Incremental cost-effectiveness ratio was estimated at £12,125 per quality-adjusted life year (QALY) gained.