BACKGROUND: Cardiovascular diseases such as acute coronary syndrome and myocardial infarction are often accompanied by severe anxiety over the likelihood of death. Cortisol has been known as a stress hormone. However, there are controversies about the effect of massage therapy on blood cortisol level. Furthermore, no study is available on the difference between massage applied by a nurse specialist or by patients’ relatives on blood cortisol level. This study was aimed to compare the effect of massage applied by a nurse specialist and patients’ relatives on blood cortisol level among the patients admitted in coronary care unit (CCU). METHODS: In a randomized controlled trial, ninety patients hospitalized at CCU were randomly placed in three groups: massage by a nurse; massage by patients’ relatives and control group. The two massage groups received a session of whole body massage. The control group received the routine care. Data were analyzed using analysis of variance, chi-square and Fischer exact tests, Kruskal–Wallis and Wilcoxon Signed Ranks tests. RESULTS: The mean age of participants was 58.43 ± 14.23 years. None of the participants had the history of massage therapy. In the group massaged by a nurse, the median blood cortisol level was 281.90 nanomoles, which were decreased to 197.00 after the intervention (P < 0.007). The median blood cortisol level in the group massaged by the patients’ relatives and the control group did not affect significantly. CONCLUSION: Massage therapy decreased the blood cortisol level in the group that received massage by a specialist nurse. It can be recommended that massage therapy be used in patients admitted in CCU.