For nearly two decades, augmented reality (AR) has found diverse applications in education, particularly in science education, where its efficacy has been supported by relevant theories and many empirical studies. However, previous studies have revealed the following research deficit: While AR technology appears to influence learning-related variables, at the time of this study only few research on the use of AR glasses in physics, a discipline for which this technology seems particularly promising in the context of laboratory experiments, has been found. Thus, the present study uses an experimental comparison group design to investigate the question of how the use of AR glasses in a physics laboratory experiment (compared to in a learning setting without AR) influences students' motivation to learn, their cognitive load during the learning process and their learning achievement. The study (sample size N = 75) investigated the impact of AR glasses in a physics laboratory experiment on optical polarization. Results align with prior research, indicating heightened motivation among learners using AR applications. However, the absence of a significant difference in cognitive load between AR and non-AR learners was unexpected. Despite expectations based on spatial contiguity, learners with AR showed no advantage in learning achievement, challenging existing meta-analyses in physics education. These findings suggest a need to shift focus from surface features, like specific AR technology, to the content design of AR applications. Future studies should analyze the deep structure of AR applications, identifying features conducive to learning.