The relation between the environment and human activity has been the subject of many empirical studies. Complexities such as which indicators related to human activity best represent biodiversity, or time lags between socio-economic activities and biodiversity, remain unclear. This paper tackles these issues based on statistical analysis of the relationship between human activity and biodiversity at the country level in Europe. We used well-adapted statistical models to explain variations in two state indicators, two pressure indicators, and one response indicator representing biodiversity. We focused on (i) the relative efficiency of the various indicators of human activity (notably human population density, human appropriation of net primary productivity, and gross domestic product per area) in predicting the biodiversity indicators, and (ii) possible time lags between metrics representing human activity and biodiversity. Results indicate that gross domestic product per area and human population density best predict the indicators for the biodiversity state, whereas human population density best predicts the indicators of biodiversity pressure. Although the indicators for biodiversity pressure best related to present-day human activity metrics, biodiversity state reveals a time lag of approximately one century. Results suggest that drivers (human population density, density of economic activity) and pressures (land sealing) should serve as primary foci for biodiversity policies. Because of the long time lags (~ one century) between these drivers and the state of biodiversity, policies regarding biodiversity should integrate a long-term view.