Many animals are exposed to periods of limited food availability at some stage of their life cycle, which could potentially lead to a feeling of hunger. To understand the animal's perception of hunger, it is essential to objectively measure subjective feelings such as hunger. This study therefore aimed to evaluate the effects of food restriction on feeding motivation in ewes. An additional aim was to develop a quantitative method for the measurement of feeding motivation as an indicator of hunger. Sixteen ewes were tested in a long, narrow race in which they could work (by walking a certain distance) to gain access to a food reward. All ewes were exposed to five different cost levels (2, 6.6, 12.1, 22, and 40 meters) and two feeding levels (24-hour restriction and an unrestricted control) for a period of 22:50 hrs per test. Data was analysed using REML and the demand function was calculated as Ln (Q) = ln(L) + b[In(P)] - a(P) with L as initial level of the demand curve, b as the initial slope and a as the acceleration of the slope with increasing cost. The results showed that 24-hour restriction significantly increased the number of rewards consumed (P<0.001) and total distance walked (P<0.001). In addition, all ewes significantly reduced the number of rewards consumed when cost was increased (P<0.001). The demand function showed that there was only a small difference in Pmax (point on the demand curve where elasticity of demand takes the value -1). However, the maximum work output at Pmax (Omax) was almost twice as large in restricted compared to unrestricted control ewes. It is concluded that 24-hour food restriction increased feeding motivation in ewes. The behavioural demand approach is therefore a suitable and sensitive measure of feeding motivation and a potential indicator of hunger in sheep.