We came to this study with a set of beliefs about good science teaching that had been heavily influenced by the constructivist literature of the past decade. In this article we reexamine some of our own assumptions about good teaching by exploring the classroom practices of an experienced physics teacher. This teacher did not fit the mold of the constructivist teacher and, yet, there was much to suggest that he was meeting the needs of the students in his class. His methods were almost entirely whole class—focusing mainly on physics content, examination technique and algorithm practice. Our close observation of this teacher in his Grade 11 classroom over several months suggests an alternative framework for examining his work. We examine this framework through a number of themes: teacher confidence, the structure of the discipline, student motivation, trust, and the cultural context of learning. We argue for a broader view of good science teaching than that proposed by the constructivist literature, one that takes into account teachers' and students' understandings of science in relation to their social and cultural contexts.