The availability of food is a crucial factor determining the health and growth of animals. Prolonged or severe food restriction will trigger the subjective state of hunger, which could potentially reduce welfare. We refined a methodology for the measurement of feeding motivation as an indicator of hunger, evaluated the effects of food restriction on feeding motivation in ewes (Ovis aries) and applied a behavioural demand analysis to the data. Sixteen ewes were exposed to five different cost levels (i.e., 2, 6.6, 12.1, 22, or 40 m walking distances per reward) and two feeding levels (24 h restriction and an un-restricted control). We found that 24 h restricted ewes worked for a higher number of rewards (P < 0.001), walked a greater total distance (P < 0.001) and approached the feeding station faster on the first visit (P < 0.001). Therefore, the methodology was able to detect differences in feeding motivation between restricted and control ewes and is a potential indirect indicator of hunger. A generic demand function fitted to the data showed that the elasticity of demand, the maximum price paid (Pmax) and maximum expenditure (Omax) were not sensitive indicators of motivation, probably because we did not generate a complete demand curve. Imposing a cost by providing sheep with the opportunity to walk for their food could be a valuable tool in establishing acceptable feeding levels from an animals’ perspective. However, more research on validating behavioural demand measures as indicators of motivation is needed.