Recent meta-analyses find various magnitudes of placebo analgesia effects in placebo mechanism trials versus placebo control trials, which have led to debate. To further investigate the magnitude of placebo analgesia in placebo mechanism trials the databases "PubMed", "PsycINFO" and "Web of Science" (2002-2007) were searched with the term "placebo analgesia". Twenty-one articles including 24 studies fulfilled the selection criteria (concerning: mechanisms, control, placebo treatment, randomization and pain measures). The validity of studies was assessed by the authors and effect sizes were calculated via difference scores. The magnitude of placebo analgesia in placebo mechanism studies was large (d=1.00) and about five times larger than placebo analgesia effects in placebo control studies (d=0.15-0.27). Differences in magnitude between the two types of studies appear to result from different types of suggestions given for pain relief. The magnitude of placebo effects was larger in studies that used long-term pain stimuli >20s (d=0.96) as opposed to short-term stimuli (d=0.81) and the largest placebo effects were found in studies wherein hyperalgesia was present (d=1.88). These results replicate our previous finding that placebo analgesic effects are higher in mechanism studies than in placebo control studies. However, since magnitudes of placebo analgesic effects are highly variable it may be valuable to investigate the factors and mechanisms that contribute to this variability as well as differences in magnitudes across types of studies.