Animal welfare measures must consider a cow’s health and production (i.e., functional state), how she feels about her situation (i.e., affective state), and whether her management allows for “natural” behaviors. Although there are accepted, quantitative measures of functional state, affective state and naturalness are typically only assessed using indirect, behavior-based measures. The most unique factors likely to affect welfare differentiating cows in grazing systems from their housed counterparts fed TMR are climatic variation, a lack of shade and shelter, and the relatively low DMI of grazing cows. Both access to and quality of feed is also less consistent and predictable in grazing cows. From a functional state perspective, fluctuating DMI is associated with risk of disease, particularly in early lactation; when affective state is considered, hunger becomes a concern as well. Although grazing cows eat less, they also produce less milk; the lack of a considerable difference in negative energy balance between unsupplemented and supplemented grazing cows in early lactation likely indicates that output is reduced to match nutrient intake. The risk of disease is minimized by ensuring that animals are adequately managed in late lactation and the dry period, are not too fat at calving, and that pre-calving energy intake is controlled. Hunger is more difficult to assess and, is often, emotively anthropomorphized. Measurements of the neuro-endocrine factors that respond to nutrient intake and expenditure have been established; however, these should be paired with behavioral measures, such as vocalizations, changes in activity and motivation to access feed. As measures are further developed for assessing general animal affective state (e.g., cognitive bias, anticipatory behavior) these should be included when assessing welfare. Although its role in the functional state of cows has been extensively investigated, further research is required to better understand the role of nutrition in cows’ quality of life in grazing systems.