The purpose of this study was to assess the efficacy of physiotherapy approaches to treatment of rotator cuff tendinopathies (RCT). Ninety shoulders were randomised in the study. A power calculation performed demonstrated using a factorial study design, 85 shoulders would be needed for 80% power at 95% confidence. All participants gave informed consent and ethical approval was granted by the Cambridge LREC. The primary outcome measure was the Shoulder Pain and Disability Index. Participants were blinded to their allocation and were randomised to one of four groups: Therapeutic Exercise(T), Manual Therapy(M), combined T/M (X), or Placebo(P). Participants were seen for two baseline assessments with a 4-week interval and then randomised. Final assessments were performed one week after the last session. The analysis involved a comparison between groups in change from baseline SPADI using ANCOVA adjusting for baseline scores. This involved testing for any interaction between M and T, and subsequently testing for main effects of M and T. Adjusted baseline and final SPADI scores (SD) by group were: X 41.6(15.4), 21.1(20.8); T 47.6(19.3), 26.3(14.7); M 44.1(17.9), 33.1(23.3); and P 39.5(24.7), 36.6(30.6). The main effects (with Significance, Standard Error and Confidence Interval) by group were: Baseline=0.686 (SE=0.104;CI=0.479,0.892); T=−13.347 (p=0.002;SE=4.091;CI=−21.479,−5.215); X=5.479 (p=0.510;SE=8.284; CI=−10.991,21.950); M=−4.126 (p=0.314;SE=4.077;CI=−12.230,3.978). A statistically significant reduction in SPADI was observed for the T group alone. There was no significant interaction effect with the addition of manual therapy and the M group did not improve significantly. It appears that best practice for treatment of RCT should centre around therapeutic exercise.