Measuring diet choice in grazing animals is challenging, complicating the assessment of feed efficiency in pasture-based systems. Furthermore, animals may modify their intake of a forage species depending on its nutritive value and on their own physiological status. Various fecal markers have been used to estimate feed intake in grazing animals. However, plant-wax markers such as n-alkanes (ALK) and long-chain alcohols may provide reliable estimates of both dietary choices and intakes. Still, their use in beef cattle has been relatively limited. The present study was designed to test the reliability of the ALK technique to estimate diet choices in beef heifers. Twenty-two Angus-cross heifers were evaluated at both post-weaning and yearling age. At each age, they were offered both red clover and fescue hay as cubes. Following 3-week acclimation periods, daily intake of each forage species was assessed daily for 10 days. During the final 5 days, fecal grab samples were collected twice daily. The ALK fecal concentrations were adjusted using recovery fractions compiled from literature. Diet composition was estimated using two statistical methods. Post-weaning, dietary choices were reliably estimated, with low residual error, regardless of the statistical approach adopted. The regression of observed on estimated red clover proportion ranged from 0.85±0.08 to 1.01±0.09 for fecal samples collected in the p.m. and for daily proportions once averaged, respectively. However, at yearling age, the estimates were less reliable. There was a tendency to overestimate the red clover proportion in diets of heifers preferring fescue, and vice versa. This was due to greater variability in ALK fecal concentrations in the yearling heifers. Overall, the ALK technique provided a reliable tool for estimating diet choice in animals fed a simple forage diet. Although further refinements in the application of this methodology are needed, plant-wax markers provide opportunities for evaluating diet composition in grazing systems in cattle.