Utilization of long-chain alcohols (LCOH) as diet composition markers in horses and cattle was assessed in a study conducted with 12 mature crossbreed mares (385 ± 47 kg BW) and six adult non-lactating cows (499 ± 36 kg BW) of Asturiana de los Valles breed. The LCOH data were combined with alkane and long-chain fatty acid (LCFA) data to test the applicability of combining these markers to estimate diet composition. Animals were randomly divided into groups of three animals and received a daily total amount of 1.0 kg dry matter/100 kg BW of diets composed of different proportions of ryegrass ( Lolium perenne) and woody species ( Ulex gallii and heather). Diet composition was estimated from even-chain LCOH (C 20 -OH to C 30 -OH) combined or not with alkane (C 25 -C 31 and C 33 ) and/or LCFA (C 22 -FA to C 28 -FA, C 30 -FA, C 32 -FA and C 34 -FA) concentrations in diet components and faeces by least-squares procedures, using marker faecal concentrations uncorrected for incomplete faecal recovery (FR0) or corrected using mean recoveries across diets within animal species (FR1). Results showed large differences between plant species in their LCOH profiles, and that these markers offered additional discriminatory information to that provided by alkanes and LCFA. The LCOH markers were incompletely recovered in the faeces of both animal species. In cattle, LCOH FR tended to increase with carbon-chain length in a linear manner in both diets ( P < 0.001), whereas in horses overall data showed a curvilinear relationship between these variables. Combination of LCOH, LCFA and alkanes resulted in more accurate diet estimates. Correction of faecal LCOH concentrations to incomplete FR led to more accurate diet composition estimates in both animal species. Results obtained in this study suggest the usefulness of LCOH markers combined with alkanes and LCFA to estimate diet composition of horses and cattle grazing mixed grassy-woody plant communities.