Although mountain dairy cattle farming systems are pivotal for the economy, as well as for social and environmental aspects. They significantly contribute to the rural development which is currently strongly prioritized in the common EU agricultural policy, at the same time they are also increasingly criticized for having relatively high environmental impacts per kg of product, such as greenhouse gas emissions. Consequently, the aim of this study was to assess and compare the environmental efficiency of 2 common alpine dairy farming systems, with focus on the effects of grazing, considering the seasonal variability in feeding at individual cow level and farm management over a 3-year period. This study focuses on alpine farming systems but can be considered to represent well other topographically disadvantaged mountain areas. An intensively managed and globally dominating production system (high-input), aiming at high milk yield through relatively intensive feeding and the use of the high-yielding dual-purpose Simmental cattle, permanently confined in stables, was compared with a forage-based production system (low-input), based on seasonal grazing and the use of the autochthonous dual-purpose breed Tyrolean Grey. For the present analysis, a data set with information on feed intake and diet composition as well as animal productivity at individual cow level, and farm management data based on multiyear data recording was used. Four impact categories were quantified for 3 consecutive years: Global Warming Potential (GWP100), Acidification Potential (AP), Marine Eutrophication Potential (MEP), and Land Use (LU, m2yr and Pt, with the latter additionally considering the Soil Quality Index). Besides being attributed to 1 kg of fat and protein corrected milk (FPCM), these impact categories were also related to 1 m2 of on-farm area. Due to limited agronomic options beyond forage production and pasture use in alpine regions, net provision of protein was calculated for both farming systems to assess food supply and quantify the respective food-feed competition. Overall, the low-input farming system had greater environmental efficiency in terms of MEP per kg FPCM, as well as MEP and AP per m2 than the high-input system. LU was found to be consistently higher for the high-input than for the low-input system, the GWP100 per kg of FPCM was lower for the high-input system. Additionally, pasture access had a significant effect on the reduction of environmental impacts. Lastly, the net protein provision was slightly negative for the high-input system and marginally positive for the low-input system, indicating a lower food-feed competition for the latter. Future studies should also address the social and economic aspects of the farming systems, to offer a comprehensive overview of the 3 key factors necessary for achieving more sustainable farming systems, particularly in disadvantaged marginal regions such as mountain areas.