Farming and natural resource extraction are the main livelihood strategies of the rural poor in developing countries. A better understanding of their relationship is needed to alleviate existing pressures on resources and to reduce poverty. To date, mainly monetary indicators have been used to measure environmental resource dependence. However, these are inadequate for poor people who consume rather than sell their environmental products. Therefore, we propose the Environmental Resource Dependence Index (ERDI) to better capture the multidimensionality of dependence. We analyse the relationship between farming efficiency and environmental resource dependence using a simultaneous equations model (SEM) and panel data for 2013, 2016 and 2017 from three rural provinces in Central Vietnam. Time-variant farming efficiency is estimated using a stochastic frontier model (SFM) with true random effects and Mundlak's adjustment. Our results show that monetary measures underestimate the extent of dependency. Therefore, policymakers should be careful to correctly identify those who are dependent on the environment. In addition, the results suggest that improved farming efficiency reduces the dependence on environmental resources. At the same time, higher dependence does not have a significant effect on farming efficiency.