The purpose of this paper is to outline some application of the Markov Process to the study of state and state changes. The essence of this mathematical concept consists of the analysis of sequences of infant responses in interaction with its environment. Categories can be defined which reflect the joint occurrence of an infant's behavior (or condition) along with some associative event(s) in the infant's immediate environment. Each of these categories of infant-environment interaction can be used as a definition of state for the purposes of studying the sequential unfolding among categories. An example utilizing child vocalization data collected by Lewis is given. When applied to mother-infant interaction, a particular mother-infant pair may yield data which give a poor fit in terms of matching statistics with the Markov model. Therefore, three alternative procedures are suggested. (Author/WY)