The purpose of this study was to examine the assumptions underlying the communicative act of interruption. Six graduate student groups involving 35 subjects were videotaped. The data for the study were 255 transcribed interruption sequences. A category system was developed and used to code the data. The results indicated that slightly over half of the interruptions served a confirming function, and the remaining interruptions were disconfirmations or rejections. No significant differences between males and females were found in the types of speeches produced. However, cross‐sex interruptions occurred significantly more often than expected.