Although pedagogical content knowledge (PCK) has become widely recognized as an essential part of the knowledge base for teaching, empirical evidence demonstrating a connection between PCK and teaching practice or student learning outcomes is mixed. In response, we argue for further attention to the measurement of dynamic (spontaneous or flexible, as opposed to static) aspects of PCK. We identify a set of trade‐offs entailed in the measurement of teachers’ PCK and propose two video‐based interview methods for eliciting teachers’ PCK. We describe a study in which we used these methods to elicit high‐school physics teachers’ PCK for the topic of force and motion. Interview 1 was based on video clips from teachers’ own classroom instruction and elicited the reasoning underlying their instructional decision making. In Interview 2, teachers responded to a standard set of video clips highlighting both typical and unexpected student thinking. Although all six teachers demonstrated the main components of their declarative PCK consistently across the two interviews, Interview 2 allowed us to characterize a more dynamic form of PCK that may underlie teachers’ in‐the‐moment instructional reasoning. When exhibiting strong dynamic PCK, teachers appeared to rely heavily on their declarative PCK as they reasoned about new examples of student thinking and corresponding instructional responses. In addition, demonstrations of dynamic PCK included features likely to support further PCK development, including a willingness to think critically about evidence of student thinking, physics content, and pros and cons of instructional representations. Although we were able to detect seemingly meaningful differences in teachers’ declarative and dynamic PCK, diversity in and the contextualized nature of the responses demonstrating strong PCK raise questions about the possibility of differentiating between such responses for measurement purposes and, thus, the potential for moving beyond elicitation, to measurement, of dynamic forms of PCK. © 2015 Wiley Periodicals, Inc. J Res Sci Teach 53: 1259–1286, 2016