This study reports the interrater agreement of the Maternal Language Classification Scale (MLCS), a functional language classification system, developed partly to avoid problems identified with previous scales. The MLCS is a comprehensive system for classifying the functional content of maternal language addressed to children whose mean length of utterance (MLU) does not exceed 3.0. Development of this scale was based on a thorough review of classification schemes used in the mother-child language literature, and on an examination of several videotape recordings of mother-child interaction to identify functional language parameters for which no categories were available. Interrater agreement for the scale was determined on a sample of 12 mother-infant dyads. Children were between 9 and 24 months of age. Audio recordings and a running narration of the stream of behavior were made while the mother and child engaged in spontaneous play in their home. Each maternal utterance was coded according to the MLCS by two independent raters. The overall agreement between raters on over 4,000 utterances was 93%. (Author/SS)