Abstract Background Exercise therapy for knee pain and osteoarthritis remains a key element of conservative treatment, recommended in clinical guidelines. Yet systematic reviews point to only modest benefits from exercise interventions. One reason for this might be that clinical trials tend to use a one-size-fits-all approach to exercise, effectively disregarding the details of their participants' clinical presentations. This uncontrolled before-after study (TargET-Knee-Pain) aims to test the principle that exercises targeted at the specific physical impairments of older adults with knee pain may be able to significantly improve those impairments. It is a first step towards testing the effectiveness of this more individually-tailored approach. Methods/Design We aim to recruit 60 participants from an existing observational cohort of community-dwelling older adults with knee pain. Participants will all have at least one of the three physical impairments of weak quadriceps, a reduced range of knee flexion and poor standing balance. Each participant will be asked to undertake a programme of exercises, targeted at their particular combination and degree of impairment(s), over the course of twelve weeks. The exercises will be taught and progressed by an experienced physiotherapist, with reference to a "menu" of agreed exercises for each of the impairments, over the course of six fortnightly home visits, alternating with six fortnightly telephone calls. Primary outcome measures will be isometric quadriceps strength, knee flexion range of motion, timed single-leg standing balance and the "Four Balance Test Scale" at 12 weeks. Key secondary outcome measures will be self-reported levels of pain, stiffness and difficulties with day-to-day functional tasks (WOMAC). Outcome measures will be taken at three time-points (baseline, six weeks and twelve weeks) by a study nurse blinded to the exercise status of the participants. Discussion This study (TargET-Knee-Pain) is the first step towards exploring whether an impairment-targeted approach to exercise prescription for older adults with knee pain may have sufficient efficacy to warrant further testing. If warranted, future randomised clinical trials may compare this approach with more traditional one-size-fits-all exercise approaches. Trial registration Current Controlled Trials ISRCTN61638364. ; RIGHTS : This article is licensed under the BioMed Central licence at http://www.biomedcentral.com/about/license which is similar to the 'Creative Commons Attribution Licence'. In brief you may : copy, distribute, and display the work; make derivative works; or make commercial use of the work - under the following conditions: the original author must be given credit; for any reuse or distribution, it must be made clear to others what the license terms of this work are.