Background: There are many claims in the horse therapy related literature of the beneficial effect of horse assisted therapy. I wanted to examine one of the most frequent claims, namely improved self-esteem. This study, therefore aimed to investigate whether a Horse Assisted Therapy intervention had any influence on the level of self-esteem in young patients undergoing treatment for substance use disorder in Norway. Design, setting and participants: This is a quantitative study which used baseline scores on Rosenberg Self-esteem scale (RSES) collected from Youth Addiction Treatment Evaluation Project (YATEP) database over a period of three years (2011-2013) from all the patients who had completed RSES once or more times (n= 119). A pretest – posttest design was used to analyze the scores of participants who had answered RSES more than once (n=54). To test the possible impact of horse assisted therapy on self-esteem, participants were divided in two groups for comparison: the intervention group (n=34) which undertook treatment as usual plus horse assisted therapy, and a control group (n=20) which undertook only treatment as usual. Additional data were extracted from the stable s schedules used for detailed registration of participation in the horse assisted therapy program, and from an electronic patient journal system which was used for validating results of treatment outcome. Findings: The participants had lower self-esteem scores when compared to average scores in the general population. They were however, comparable with score for a similar group of substance use disorder patients in the USA. There was no significant difference in baseline self-esteem (n=119) for the intervention group and the control group (p=0.066). Both groups displayed a low level of self-esteem compared to a normal Norwegian population. There was no significant increase of self-esteem in ether groups during the study. I found no significant difference arising from horse assisted therapy between the intervention group and the control group. While I found no empirical evidence of self-esteem being a predictor of treatment outcome as a whole, I did find, post hoc, that higher self-esteem was a predictor of positive treatment outcome for males but not females. Conclusion: I found no evidence to support claims that horse assisted therapy facilitates improved self-esteem. There was, however, a possible indication that horse assisted therapy may provide a means of attracting young addicts, particularly females, with a low self-esteem into therapeutic treatment. This requires further investigation.