Advances in animal biologging technologies have greatly improved our understanding of animal movement and distribution, particularly for highly mobile species that travel across vast spatial scales. Assessing the accuracy of these devices is critical to drawing appropriate conclusions from resulting data. While understanding the vertical dimension of movements is key to assessing habitat use and behavior in aerial species, previous studies have primarily focused on assessing the accuracy of biologging devices in the horizontal plane with far less emphasis placed on the vertical plane. Here we use an Unaccompanied Aircraft System (UAS) outfitted with a laser altimeter to broadly assess the accuracy of altitude estimates of three commonly used avian biologging devices during three field trials: stationary flights, continuous horizontal movements, and continuous vertical movements. We found that the device measuring barometric pressure consistently provided the most accurate altitude estimates (mean error of 1.57m) and effectively captured finer-scale vertical movements. Conversely, devices that relied upon GPS triangulation to estimate altitude typically overestimated altitude during horizontal movements (mean error of 6.5m or 40.96m) and underestimated amplitude during vertical movements. Additional factors thought to impact device accuracy, including Horizontal- and Position- Dilution of Precision and the time intervals over which altitude estimates were assessed, did not have notable effects on results in our analyses. Reported accuracy values for different devices may be useful in future studies of aerial species' behavior relative to vertical obstacles such as wind turbines. Our results suggest that studies seeking to quantify altitude of aerial species should prioritize pressure-based measurements, which provide sufficient resolution for examining broad and some fine-scale behaviors. This work highlights the importance of considering and accounting for error in altitude measurements during avian studies relative to the scale of data needed to address particular scientific questions.