The aim of this study was to determine whether horses exhibiting unilateral hindlimb lameness unload (rest) the lame limb more than the contralateral limb. The resting/unloading of the hindlimbs and the time spent lying down were measured using accelerometers. Ten non-lame horses and 20 lame horses were recruited for participation and monitored for 11 h overnight with accelerometers (MSR145, sampling rate: 1 Hz, and measuring range: ±15 g) attached to the lateral metatarsal and metacarpal regions of each limb. Metatarsal and metacarpal orientation were used to determine whether the limb was unloaded (rested) or loaded, respectively, or whether the horses were lying down. The relation of resting time between non-lame and lame limbs (non-lame/lame: 0.85 ± 1.2) of the lame horses differed significantly (p = 0.035) from that of the non-lame horses (right/left: 1.08 ± 0.47). Non-lame horses rested their hindlimbs evenly (left: 15 ± 10%; right: 17 ± 16%). Horses with unilateral hindlimb lameness unloaded the lame limb longer (lame limb: 61.8 ± 25.3%, non-lame limb: 38.2 ± 25.3%) than their contralateral limb. The lame horses (13 ± 11%) lay down longer (p = 0.012) than the non-lame horses (3 ± 6%). The degree of lameness determined by the participating veterinarians (Vet Score) (r = −0.691, p < 0.01) and the asymmetry evaluated by the lameness locator (ALL) (r = −0.426, p = 0.019) correlated with the resting ratio (rest time ratio). Both factors were also correlated with the time spent lying down (Vet Score (r = 0.364, p = 0.048) and the ALL (r = 0.398, p = 0.03)). The ALL and VET Score were significantly correlated (r = 0.557, p = 0.01). The results of this study provide a good baseline for future research into how individual resting patterns may help to detect pain.