Biodiversity offsets are increasingly employed as an approach to compensate for unavoidable development impacts. Reliance on overly simplistic metrics in assessing the impacts of development, and assigning offset requirements, generally results in offsets which fail to conserve the key ecological values they seek to protect. We conducted a cross-disciplinary quantitative review, based on 255 peer-reviewed publications from three fields of research; offsetting (n = 43), conservation planning (n = 54) and ecology (n = 158), to explore which metrics are commonly used in offsetting compared to the conservation and ecology literature. We recorded the use of biodiversity metrics from 24 categories which captured broad habitat patterns (e.g. habitat area and condition) as well as specific biological and ecological mechanisms (e.g. diversity, population density or landscape connectivity). Our review found that offset studies and programs rely heavily on habitat attributes and area-based metrics, with >70% of the offset literature having used these metrics. Habitat attributes and area-based metrics were less frequently reported in the conservation planning (56 and 59%, respectively) and ecological literature (49 and 15%). Ecological research had a higher frequency of metrics reflecting the biological and ecological processes relevant to biodiversity, such as species’ population densities and species-specific connectivity. Our results also indicate a notable disconnect in how biodiversity is measured when offsets are planned compared to when their outcomes are evaluated. This demonstrates the need to re-evaluate the way offset policies and programs value, describe and measure biodiversity, so that critical biodiversity values and important ecological processes are appropriately captured, and no net loss is achieved.