Freezing is an adaptive defensive response to a stressful event. Recent research suggests that freezing not only occurs in response to physical threats but also in response to social threats (e.g., angry faces; Roelofs et al. in Psychol Sci 21:1575-1581, 2010). Given the practical and theoretical importance of this finding, the current study aimed to replicate and extend it. Following the original study, we measured heart rate while participants viewed emotional faces (angry, happy, neutral). Extending the original study, we included a baseline measure and performed additional, more fine-grained analyses. Our results support the hypothesis that participants show physiological signs of freezing when looking at angry faces. Importantly, we also find this effect when comparing heart rate in the angry block to baseline levels. Interestingly, the heart rate effects are explained by deceleration in the first 30 s of the 1-min angry block, but not in the second 30 s. Like Roelofs et al., we find evidence that the effects are modulated by state anxiety, but our effects are only marginal and we do not replicate the negative correlation between heart rate and state anxiety in the angry block. In general, we thus find evidence for physiological signs of freezing in response to social threat. We discuss implications and venues for future research.