This paper reports a case study of four grade 10 physical sciences teachers' PCK about graphs of motion. We used three data collection strategies, namely teachers' written accounts, captured by the content representation (CoRe) tool, interviews and classroom observations. We conceptualised the PCK displayed in the CoRe tool and the interviews as "reported PCK" and the PCK demonstrated during teaching as "enacted PCK." These two manifestations of PCK were compared to establish the extent of agreement between reported and enacted PCK. We adopted the topic-specific PCK (TSPCK) model as the framework that guided this study. This model describes TSPCK in terms of five components of teacher knowledge. Guided by the model, we designed two rubrics to assess these manifestations of TSPCK on a four-point scale. The results of this study indicated that the reported PCK was not necessarily a reflection of the PCK enacted during teaching. The levels of PCK in the components were seldom higher in the enacted PCK, but tended to be similar or lower than in the reported PCK. The study implies that the enactment of PCK should be emphasised in teacher education.