Objective To assess the clinical effectiveness and cost-effectiveness of glucosamine sulphate/hydrochloride and chondroitin sulphate in modifying the progression of osteoarthritis (OA) of the knee. Data sources Electronic databases were searched from 1950 to 2008 and included: MEDLINE and PubMed; EMBASE; Cochrane Library (including Cochrane Systematic Reviews Database, CENTRAL, DARE, NHS EED and HTA databases); Allied and Complementary Medicine (AMED); National Research Register (NRR); Web of Science Proceedings; Current Controlled Trials; and Clinical Trials.gov. Other sources included bibliographies of retrieved papers, registered but unpublished trials, internet searches and the Food Standards Agency website. Review methods A search was conducted for systematic reviews of randomised controlled trials (RCTs), which were used to identify RCTs of at least 12 months’ duration and updated with searches for primary studies. A cost-effectiveness model was constructed using cohort simulation and drawing on available evidence. Sensitivity analysis was undertaken and value of information analysis conducted. A review of studies of mechanism of action was carried out to explore the biological plausibility of the preparations. Results Five systematic reviews and one clinical guideline met the inclusion criteria. They reported inconsistent conclusions with only modest effects on reported pain and function. A reduction in joint space narrowing was more consistently observed, but the effect size was small and the clinical significance uncertain. A separate review of eight primary trials of > 12 months’ duration showed evidence of statistically significant improvements in joint space loss, pain and function for glucosamine sulphate, but the clinical importance of these differences was not clear. In two studies of glucosamine sulphate, the need for knee arthroplasty was reduced from 14.5% to 6.3% at 8 years’ follow-up. For other preparations of glucosamine, chondroitin and combination therapy, there was less evidence to support a clinical effect. Cost-effectiveness modelling was restricted to glucosamine sulphate. Over a lifetime horizon the incremental cost per quality-adjusted life-year (QALY) gain for adding glucosamine sulphate to current care was estimated to be £21,335. Deterministic sensitivity analysis suggested that the cost-effectiveness of glucosamine sulphate therapy was particularly dependent on the magnitude of the quality of life (QoL) gain, the change in knee arthroplasty probability with therapy and the discount rate. At a cost per QALY gained threshold of £20,000, the likelihood that glucosamine sulphate is more cost-effective than current care is 0.43, while at a threshold of £30,000, the probability rises to 0.73. Probabilistic sensitivity analysis showed that estimates were imprecise and subject to a degree of decision uncertainty. Value of information analysis demonstrated the need for further research. Several biologically plausible mechanisms of action for glucosamine sulphate and chondroitin were proposed. Conclusions There was evidence that glucosamine sulphate shows some clinical effectiveness in the treatment of OA of the knee. No trial data came from the UK and caution should be exercised in generalising the findings to the UK health-care setting. Cost-effectiveness was not conclusively demonstrated. There was evidence to support the potential clinical impact of glucosamine sulphate. The value of information analysis identified three research priorities: QoL, structural outcomes and knee arthroplasty. The biological mechanism of glucosamine sulphate and chondroitin remains uncertain and, in particular …