OBJECTIVE: The purpose of this narrative review was to highlight recent research in the rehabilitation of people with osteoarthritis (OA) by summarizing findings from selected key systematic reviews and randomized controlled trials (RCTs). METHODS: A systematic search was conducted using the PubMed, Physiotherapy Evidence Database (PEDro) and Cochrane databases from April 1st 2014 to March 31st 2015. A selection of these is discussed based on study quality, relevance, contribution to new knowledge or controversial findings. Methodological quality of RCTs was assessed using guidelines from PEDro. RESULTS: From 274 articles, 74 were deemed to meet the eligibility criteria including 24 systematic reviews and 50 studies reporting on findings from RCTs. Overall the methodological quality of the RCTs was moderate. The studies were grouped into several themes covering; evidence of rehabilitation outcomes in less studied joints including the hand and hip; new insights into exercise in knee OA; effects of biomechanical treatments on symptoms and structure in knee OA; and effects of acupuncture. CONCLUSIONS: Exercise was the most common treatment evaluated. Although little evidence supported benefit of exercise for hand OA, exercise has positive effects for hip and knee OA symptoms and these benefits may depend upon patient phenotypes. The first evidence that a brace can influence knee joint structure emerged. The latest evidence suggests that acupuncture has, at best, small treatment effects on knee OA pain of unlikely clinical relevance.