A randomized controlled study was implemented on 90 nursing students in Turkey. The sample was divided into two groups: intervention and control group. The intervention group received education on refugee health and participated in a practical internship with refugees. Three scales were administered to the two groups at baseline (pre-test) and the end of intervention (post-test): "Xenophobia", “Attitude Towards Refugees” and "Intercultural Sensitivity". The three scale post-test scores of the intervention statistically significantly differed from control group and showed a statistically significant interaction between the groups and time in the two-way analysis of variance. The three scale post-test scores significantly differed from the pre-test scores in the intervention group, showing more relevant effects on xenophobia scale, two subscales of attitude towards refugees and intercultural sensitivity scale. The interventions based on the contact hypothesis improved nursing students' attitudes towards refugees.