In this paper, we report work undertaken with a group of 11 UK teachers over a period of a year to teach aspects of the nature of science, its process, and its practices. The teachers, who taught science in a mix of elementary, junior high, and high schools, were asked to teach a set of “ideas-about-scienc” for which consensual support had been established using a Delphi study in the first phase of the project. Data were collected through field notes, videos of the teachers' lessons, teachers' reflective diaries, and instruments that measured their understanding of the nature of science and their views on the role and value of discussion in the classroom. In this paper, drawing on a sample of the data we explore the factors that afforded or inhibited the teachers' pedagogic performance in this domain. Using these data, we argue that there are five critical dimensions that distinguish and determine a teacher's ability to teach effectively about science. Whilst these dimensions are neither mutually independent nor equally important, they serve as a valuable analytical tool for evaluating and explaining the success, or otherwise, that individual teachers of science have when confronted with teaching aspects about science. In addition, we argue that they are an important means of identifying salient aspects of pedagogy for initial and in-service training of science teachers for curricula that incorporate elements of “ideas-about-science” © 2004 Wiley Periodicals, Inc. Sci Ed, 88:655–682, 2004