This study examined whether placebo responses were predicted by a theoretical model of specific and general treatment beliefs. Using a randomised cross-over, experimental design (168 healthy individuals) we assessed whether responses to a cold pressor task were influenced by two placebo creams described as Pharmaceutical vs Natural origin. We assessed whether placebo responses were predicted by pre-treatment beliefs about the treatments (placebo) and by beliefs about the pain. The efficacy of both Pharmaceutical and Natural Placebos in reducing Pain Intensity was predicted by aspects of pain catastrophizing including Feelings of Helplessness (Pharmaceutical: B=0.03, p<0.01, Natural: B=0.02, p<0.05) and Magnification of Pain (Pharmaceutical: B=0.04, p<0.05, Natural: B=0.05, p<0.05) but also by pre-treatment Necessity beliefs (Pharmaceutical: B=0.21, p<0.01, Natural: B=0.16, p<0.05) and, for the Pharmaceutical condition, by more general beliefs in personal sensitivity to pharmaceuticals (B=0.14, p<0.05). Treatment Necessity beliefs also partially mediated the effects of Helplessness on placebo responses. Treatment Necessity beliefs for the Pharmaceutical Placebo were influenced by general pharmaceutical beliefs whereas Necessity beliefs for the Natural Placebo were informed by general background beliefs about holistic treatments. Our findings demonstrate that treatment beliefs influence the placebo effect suggesting that they may offer an additional approach for understanding the placebo effect.