This research assessed the relative impact of habit and behavioral intentions in predicting classroom teacher behavior, using a model proposed by Triandis. Responses from a behavioral differential, as well as two hours of classroom observations, were taken on 77 male and female black and white junior high school teachers. The classroom observation technique (STOIC) obtained the frequencies of emitted behaviors (both verbal and nonverbal), categorized by race and sex of the target child. Results indicated that habit was a more potent predictor of classroom behavior than intentions. However, a post-hoc analysis supported the notion that intentions become important when the habit component can be suppressed.