This paper reexamines the claims about the historical source of nasalized vowels in languages. The author makes use of data from Òko, a West-Benue Congo language, to show that nasalized vowels evolved from a vN sequence (Greenberg 1966), and not CNV like some scholars have claimed for some African languages (Hyman 1972; Williamson 1973). The facts about nasalization in Òko are stated as follows: nasalized vowels are oral vowels in the environment of a syllable-final N; the syllable-final N only gets a surface realization across a morpheme boundary; the surface form of the syllable-final N alternates between [n] and [m]. The choice of any of the variants of the syllable-final N is phonologically predictable. That is, [n] is triggered by [-back] vowels, while [m] is triggered by [+back] vowels. However, the rule is not as straightforward as described. Under certain phonological conditions, there is an overlap in the distribution of the variants of the N-coda. This is as a result of the language's effort to prohibit the co-occurrence of adjacent identical segments at the suprasegmental level.