It is widely acknowledged that we are in the midst of an extinction crisis and habitat loss is generally considered the primary driver. However, providing accurate estimates of extinction rates has proven to be problematic and a range of extinction estimates have been published. Arguably, the most commonly used method for predicting extinctions resulting from habitat loss has been application of the species–area relationship (SAR). The purpose of this chapter is to provide a review of the many ways in which the SAR has been used to predict the number of extinctions resulting from habitat loss. By doing so, we highlight the pitfalls of using the SAR in such a way and discuss how the SAR has been argued to both over-predict and under-predict extinctions. We also provide examples of the myriad ways in which studies have extended and built on standard SAR models and approaches to better model and predict extinctions. We conclude by arguing that there is a need to recognize that any approach based on a single variable (i.e. area), such as the SAR, is unlikely to provide a perfect extinction prediction, regardless of the specific details.