Development resilience is defined as the sustained capacity of an entity to avoid falling below a normative threshold of well-being. This paper extend this concept and measurement method to jointly model the resilience of both herding communities and rangelands using seemingly unrelated regression (SUR) to account for the reciprocal feedback effects and common external shocks they face. Using high-quality and high-frequency disaggregated spatial data from Mongolia, this paper estimates the community-level conditional probability of attaining minimal threshold values for herdsize per capita and rangeland productivity (measured by NDVI) as the main outcome variables. The findings reveal that herding community resilience is primarily influenced by past livestock holdings and accumulated precipitation, with extreme hot and cold temperatures reducing resilience. Rangeland resilience, on the other hand, is mostly associated with stocking rate, with increased grazing pressure negatively impacting resilience. While summer precipitation shows a weak positive association with rangeland resilience, extreme cold days in winter have a significant negative association, and extreme hot days do not significantly affect rangeland resilience. The paper contributes to the development resilience literature by providing the first empirical study of joint resilience of a coupled human- nature system. The findings have implications for policy and future research, emphasizing the need for sustainable grazing management and the potential for simultaneously studying the causal impacts of development programs and climate change on human well-being and resource health.