English education policy-makers have targeted classroom time as a key area for regulation and intervention, with "brisk pace" widely accepted as a feature of good teaching practice. We problematise this conventional wisdom through an exploration of objective and subjective dimensions of lesson pace in a corpus of 30 Key Stage 2 literacy lessons from three classrooms in one London school. Systematic classroom observation produced an anomaly: the lessons we experienced as fast-paced were rated objectively as slowest, and vice-versa. We contrasted the fastest and slowest episodes in the corpus, demonstrating that for these episodes the accepted measure of pace primarily reflected differences in utterance length. Linguistic ethnographic microanalysis of the episodes highlighted predictability, stakes, meaning and dramatic performance as key factors contributing to pace as experienced. We argue, among other claims, that sometimes accelerating pupils' experience--and learning--necessitates slowing down the pace of teaching, and that government calls for urgency may, perversely, make lessons slower. (Contains 2 tables, 2 figures, and 17 notes.)