The nature and extent of code-mixing in the language acquisition process over a 12-month period was studied with a child growing up in a bilingual English-Chichewa (Bantu language) speaking home. Data are examined from age 18 1/2 months to 30 1/2 months. Definitions of code-switching are offered, and an analysis of code-mixing in the speech of the author's daughter is presented. Utterances were abstracted that contained: (1) loan words and loan blends that are in common usage in Chichewa speakers; (2) the family's loan words for which Chichewa has no equivalent; (3) genuine mixing, where both languages had different words, but the child either did not know the word in the language she was speaking, or did not use it for that particular utterance; (4) morphological borrowing; and (5) examples where the child seemed to be aware of mixing, and switched to the language in which she could say the whole sentence. The paper includes speech samples, data on percentages of types of mixing, data on percentages of syntactic utterances mixed, and a list of utterances that were mixed at the level of the phrase. (SW)