Genetic analysis of non-invasively obtained samples is an increasingly affordable option for many wildlife studies, but it has remained difficult to obtain high-quality samples from many species. We modified 8” Belisle foot snares (Belisle Enterprises, Quebec, Canada) to non-invasively obtain mountain lion (Puma concolor) hair samples in unbaited trail sets. We deployed 22 hair traps, monitored by remote cameras, at 66 locations for 1618 active trap nights (x̄= 24.5 nights, SD = 7.2 nights). Photos indicated 20 instances of mountain lions passing within 2 m of a hair trap and we collected 7 mountain lion hair samples, which averaged >20 hairs/sample. All samples contained hair with visible roots and were identifiable to species; 6 of the 7 (85.7%) yielded sufficient DNA for individual identification. We attributed failure to obtain samples to 3 primary causes: individual trap saturation (2 instances), trap failure (2 instances), and non-trigger events (9 instances). Black bears (Ursus americanus) and heavy rains were the primary sources of disturbance to hair trap sets, contributing to individual trap saturation and trap failure. We speculate that low trigger rates were associated with pan tension having been set too high in the first month of the study, as well as disturbance of hair traps or leading foot placements by nontarget species. We discuss strategies to increase hair sample collection rates, including seasonal use of hair traps, more selective placement on the landscape, and altering physical attributes of the hair traps. Taking these strategies and the quality of hair samples collected into account, we believe hair traps are a viable tool for noninvasively collecting genetic material for individual identification of mountain lions and other elusive species. These data can be applied to studies of habitat connectivity, breeding success and relatedness, population density, metapopulation structure, or any others in which a bank of individual genotypes are useful.