This research discusses the types and functions of code mixing found in the "Indonesia's Next Top Model Cycle 3" event. This research focuses on analyzing the code mixing used by the models and judges at the event. The main objective of this research is to find out the types and functions of code mixing at the "Indonesia's Next Top Model Cycle 3" event, then find out the types of code mixing that are most widely used, analyze the reasons and functions of certain code mixing. The data for this research are 62 clauses containing code mixing at the event "Indonesia's Next Top Model Cycle 3". This research was conducted using Qualitative Analysis Research. The technique used to collect data is the technique of recording, reading, studying and analyzing all the data. The data in this research were analyzed using Hoffman's code mixing theory (1991). Based on data analysis, it was found that there were three types of code mixing found at the "Indonesia's Next Top Model Cycle 3" event, namely intra-sentential code mixing, intra-lexical code mixing, and involving pronunciation changes. furthermore, there are seven functions of code mixing, namely talking about certain topics, quoting other people, being empathetic towards something, interjection, repetition used for clarification, intention to clarify the content of the interlocutor's speech, and expressing group identity. The most dominant type of code mixing used in the "Indonesia's Next Top Model Cycle 3" event was Intra Sentential Code Mixing with a total of 57 clauses out of the 62 clauses found. Furthermore, the most dominant function of code mixing is found in the "Indonesia's Next Top Model Cycle 3" event, namely discussing certain topics, namely there are 39 clauses out of 62 clauses.