Code-mixing involves the deliberate mixing of two languages without an associated topic change. It is primarily used as a solidarity marker. It is not something brought about by laziness or ignorance as such, rather it requires the conversant to have a good knowledge of the grammar of the two languages and to be well aware of societal norms. It is a source of pride to bilinguals (Wardhaugh 1986). In this paper we examine code mixing from the interference dimension, looking at it from the phonological and inter-lingual angles. We discuss code-mixing because Efik people are not monolingual. To them, substantial command of English is a passport to the arena of globalization and competitive white-collar job market. Therefore mixing Efik and English is inevitable. Urbanization, education, government business and multilingualism have triggered the Efik people to learn English. A combination of research principles using unstructured forms of data collection research methods is used for this study which are (i) Participant Observation and (ii) In-depth Interview. This paper is rooted in the phonemic theory which models what happens to the languages when there is a mixing and interference. We used aspects of morphological and sociolinguistic models in the analysis. We have come up with the key findings which state that the grammatical items, rather than the lexical ones, are crucial to the identity of a language. Also a language may borrow lexical items freely assimilating or not assimilating them. We can again add that a language is on its way to losing its identity once it starts borrowing grammatical items from another language.