In the French ruminant feeding system, the equation for estimating intake capacity (IC) for sheep (expressed as fill unit, UEM) makes it possible to predict the voluntary dry matter intake (DMI). The equation for sheep at maintenance was established some 40 years ago on the suckling sheep breed, never updated, and applied to the dairy breed. The present experiment was designed to measure the IC of Lacaune dairy ewes. Two experiments were conducted in winter and early spring (Exp1) and autumn (Exp2) with dry and non-pregnant adult Lacaune dairy ewes. Thirty-six ewes (79 ± 8.3 kg) in Exp1 and 24 ewes (74 ± 7.6 kg BCS) in Exp2 were separated into two batches of two groups (Exp.1: G1L, G1H, G2L, G2H; Exp. 2: G3L, G3H, G4L, G4H) based on low (L) and high (H) body weight (BW) with a difference ranging from 9 to 14 kg. Feed was exclusively limited to three hays (H1, H2, H3) in Exp1 and two (H4, H5) in Exp2. Hays H1 and H5 had a low nutritive value (5–6% crude protein, CP) with high fill value (1.76 UEM, low intake), whereas the others had medium nutritive value (8–12% CP) with medium fill value (1.34–1.56 UEM). In Exp1, in winter, G1 received H1 hay and G2 received H2 hay for two weeks; the hays were then reversed for the following two weeks. In spring, G1 received H3 hay and G2 received H2 hay for two weeks only. In Exp2, G3 received H4 hay and G4 received H5 hay for two weeks, and the hays were then changed for the following two weeks. The DMI was measured every week for four days, and BW and BCS were measured every two weeks. The IC observed (DMI times the fill value of the hay) was calculated per group and week. Surprisingly, DMI was not dependent on BW (P > 0.05) except in two cases, which may have been due to a high average BCS (3.2 ± 0.25), which is known to be linked to high plasmatic leptin levels that have a negative impact on DMI. When expressed as metabolic BW, DMI was higher (P < 0.05) in autumn than in winter. In all cases, the IC observed was always higher (P < 0.00001) than the IC calculated with the present equation by approximately 0.8 UEM, representing a difference of 0.44–0.62 kg DMI depending on the hay quality. In conclusion, the present equation for estimating the IC of Lacaune dairy ewes underestimates DMI. More studies, including various BCSs, must be conducted to confirm these results for establishing a new equation for Lacaune dairy ewes and for other breeds as well.