The aim of this research is to investigate code mixing produced by the students of ‘Debate Activity as part of English Club’ program at SMAN 2 Bengkulu city. The design are descriptive quantitative and qualitative. The subject is eight students who are involved in the debate activity. The instruments consist of observation checklist and interview. The data are taken from the recording of the debate activity on 18th May 2018. In analyzing the data, the researcher trancribes the recording, classifies the data which can be categorized as code mixing based on Fasold’s theory (2004), analyzed the reasons why the students mixed the code, made a percentage and draw conclusion. The result showed that there were two types of code mixing applied by the students in debate activities. The type of code mixing appeared were insertion of word code mixing (56.6%) and insertion of phrase code mixing (43.3%). The data of interview were some reasons why the students mixed the code in the debate activity, namely repetead used for correction, because of unintentionally, and because of less vocabulary. However, the code mixing could not be denied in debate activity as part of English club programs.