It is of practical significance to clarify farmers' ecosystem service dependence and its multi-level influencing factors from the perspective of rural revitalization for local rural revitalization practices and farmers' well-being improvement. Taking Mizhi County of Shaanxi Province in the ecologically fragile area as study region, the farmers' ecosystem service dependence was evaluated based on the value method. At the same time, the differences between villages and between the farmers in the index of dependence on ecosystem services were analyzed from the perspective of rural revitalization, and finally the multi-level model was used to reveal the influencing factors of the farmers' index of dependence on ecosystem services. The results were as follows: (1) the average value of the farmers' index of dependence on ecosystem services was 0.673. The local farmers mainly relied on the benefits of provisioning services, and were less dependent on regulating services and cultural services. The local farmers mainly depended on the benefits of crops in provisioning services, and their dependence on fruits and livestock was relatively low. (2) The farmers in rural revitalization demonstration villages were more dependent on ecosystem services than those in non-demonstration villages, and there was strong village heterogeneity in the various ecosystem services dependence. The proportion of farmers in the ecosystem service combination types of "crops plus livestock plus regulating services", "crops plus fruits plus regulating services", and "crops plus regulating services" accounting for about 80% of total farmers, and the indexes of dependence on ecosystem services were 0.690, 0.835, 0.410, respectively. The diversity index of farmers was 2.882, and the diversity indexes of farmers in demonstration villages were higher than that in non-demonstration villages. The more diverse the ecosystem service types of the farmers, the higher the farmers' index of dependence on ecosystem services. (3) Individual variables such as age, proportion of patients, per capita annual income, and per capita arable land affected the index of dependence on ecosystem services. Among the environmental variables, villages with higher soil conservation ability had higher dependence on crops and livestock; villages with more government funds had higher dependence on fruits and cultural services. At the same time, altitude was significantly positively correlated with fruit dependence, and population density was significantly negatively correlated with regulation services.