Feedback of medical imaging results can reveal visual evidence of actual bodily harm attributable to a given behaviour. This may offer a particularly promising approach to motivating changes in health behaviour to decrease risk. Applicable behaviours include smoking cessation, skin self-examination, sun protection behaviour, dietary intake, physical activity and medication usage. The current review assembles and evaluates the evidence concerning the behavioural impact of showing and explaining images, in order to determine whether their communication is an effective intervention approach. To assess the extent to which feedback to individuals of images of their own bodies created during medical imaging procedures increases or decreases a range of health behaviours. We searched the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 3 2009), MEDLINE (1950 to 14 September 2009), EMBASE (1980 to 14 September 2009), CINAHL (1982 to 9 October 2009), PsycINFO (1806 to 14 September 2009) and reference lists of articles. We also contacted authors of selected papers, and searched the ProQuest Dissertations and Theses database on 1 October 2009 for grey literature. Randomised or quasi-randomised controlled trials involving adult (18 years and over) non-pregnant individuals undergoing medical imaging procedures assessing risk of disease or of an existing condition, for which personal risk may be reduced by modification of behaviour. The sole or principal component of included interventions is visual feedback of individuals' medical imaging results, defined as individuals being shown, and having explained, source images (still or moving images) of their bodies generated by the procedure. Two authors searched for studies and independently extracted data from included studies, with disagreements resolved by consensus and a third author acting as arbiter. The risk of bias of included studies was assessed and reported in accordance with the guidelines in the Cochrane Handbook for Systematic Reviews of Interventions. We conducted a narrative synthesis of the included studies, dividing them into clinical and non-clinical population groups and presenting major characteristics and results. Where the studies were sufficiently similar in terms of population, inclusion criteria, interventions and/or outcomes, we pooled the data statistically. We included nine trials involving 1371 participants. Overall, results were mixed. Regarding five trials in clinical populations, three assessed smoking cessation behaviours, all featuring arterial scanning procedures to assess cardiovascular risk, and reported a statistically significant effect favouring the intervention, producing a pooled odds ratio (OR) of 2.81 (95% confidence interval (CI) 1.23 to 6.41, P = 0.01). One of these trials also measured physical activity and reported no statistically significant difference between the groups. A further trial measured skin examination behaviour following a skin photography procedure for assessing moles, and reported a statistically significant increase in favour of the intervention, with an OR of 4.86 (95% CI 1.95 to 12.10, P = 0.0007). The final clinical population trial measured a range of dietary intake and medication usage behaviours and featured an arterial scanning procedure assessing cardiovascular risk, and reported no statistically significant effects.Among the four trials in non-clinical populations, all featuring ultraviolet (UV) photography to highlight UV-related skin damage, a statistically significant result favouring the intervention was found in one trial for reducing tanning booth use, producing a mean difference (MD) of -1.10 (95% CI -1.90 to -0.30, P = .007) and one trial reported an outcome on which the control condition was favoured, with an MD of 0.45 (95% CI 0.04 to 0.86, P = 0.03) on intentional hours spent in the sun. In two further trials, no statistically significant behavioral effects were reported regarding time spent in the sun or sun protection behaviours.There was no evidence of significant adverse effects in the included trials, although this was not well reported. Due to the limited nature of the available evidence and the mixed results that were found, no strong statements can be made about the effectiveness of communicating medical imaging results to change health behaviour. Only three trials in clinical populations were similar enough in term of setting, intervention and outcome to allow meta-analysis. We suggest, however, that targeted interventions using medical imaging technologies may be effective in certain contexts, or as applied to certain behaviours, but that this should be considered on an intervention by intervention basis, and not assumed as a general principle.