Music is considered an integral part of everyday life for most humans. Over the past decades, it has emerged as a therapeutic tool, widely used in medical interventions. It is valued for being cost-effective and associated with minimal side effects. Several meta-analyses have demonstrated beneficial effects of music therapy on various pathological conditions, including dementia (Lin et al., 2023), blood pressure (do Amaral et al., 2016), and chronic pain (Garza-Villarreal et al., 2017). However, music is a complex stimulus consisting of different core elements that define the character of a piece, such as melody, rhythm, harmony, or dynamics (Tzanetakis and Cook, 2002; Large et al., 2023; McFee et al., 2015). Therefore, music therapy in humans is often individualized according to the preferences of the patient, including the genre of music, duration of exposure, and sound pressure level (American Music Therapy Association, n.d.; de Witte et al., 2022). The perceived favorability of music was even found to be critical for its analgesic effect, irrespective of musical genre (Van der Valk Bouman et al., 2024). The multidimensional character of music poses a challenge for standardizing music interventions in both humans and animals. The same musical piece might produce significantly different effects across individuals, with an individual’s current affective state further influencing outcomes. While music interventions studies in humans can be tailored to the preference of the experimental subject, this is not possible in animal studies, making it essential to evaluate music preference and perception first.

In general, animal models are valuable tools in experimental research, as they enable detailed analysis of neurobiological mechanisms which cannot be directly studied in humans. To complement positive findings from human studies, the effects of music have been examined across various animal models to provide a multilevel perspective, integrating behavioral, physiological, and structural analyses. This approach allows researchers to explore how music influences brain structure, behavior, and homeostasis in both healthy animals and disease models, providing insights into translational applications (Kühlmann et al., 2018).

Numerous studies have shown a positive influence of music interventions on rodents, notably, enhanced neurogenesis and neuroplasticity (Lee et al., 2016; Fu et al., 2025; Xing et al., 2016b), improved learning ability (Xing et al., 2016a; Trzesniak et al., 2024), or a positive effect on blood pressure (Akiyama and Sutoo, 2011). Most of the studies examining music interventions in rodents used classical music as a stimulus (Kühlmann et al., 2018), especially the Sonata for Two Pianos in D major, K.448 by Mozart is regularly used as an auditory stimulus in such experiments. Listening to this masterpiece was originally reported to enhance short-term, temporary spatial-reasoning in humans (Rauscher et al., 1993). The deduced so-called “Mozart effect” is nowadays considered to be more of a myth than an actual effect (Oberleiter and Pietschnig, 2023). Beyond that, rodents and especially mice have a hearing range that notably differs from humans (Heffner and Heffner, 2007). Rats can hear frequencies from around 500 Hz up to 64 kHz, while mice have a hearing range of about 2 kHz to 100 kHz. In comparison, the human auditory range spans from 20 Hz to 20 kHz (Heffner and Heffner, 2007). Most of Sonata K.448 is below 1 kHz, only some tunes go up to 3 kHz. Thus, it is unlikely that mice can perceive the composition in the same manner as humans do. Adding to this problem, music preference is based on ethnological and individual factors.

While the importance of studying the effect of music on animals is indisputable, the type of music used in such studies needs to be carefully adjusted to the respective species.

Yet, the key parameters of music interventions are difficult to assess in rodents, given that the musical preference of mice is still not well understood within the scientific community.

In this pilot study, we aimed to develop a method capable of determining music preferences of wild-type C57BL/6J mice given a choice between four distinct genres: pop, electronic dance music (EDM), rock, and classical music.

Our study unexpectedly reveals that classical music, represented by Sonata K.448 by Mozart, was associated with less time spent in the respective chamber when directly compared to the other presented compositions.

To date, the scientific community lacks information regarding the music preference of mice. Despite this, music is recurrently used in murine therapeutic interventions, often without considering how mice actually perceive it. This highlights the need for novel approaches to understand how auditory stimuli affect mice.

Our findings in the music genre test indicate that wildtype C57BL/6J mice spent comparable amounts of time in chambers playing EDM, Rock and Taylor Swift, while showing reduced occupancy in the compartment presenting Sonata K.448. This behavioral response may suggest a relative avoidance of Sonata K.448 when presented as an alternative, although definite statements on avoidance and preference require further testing, particularly using silence as a primary negative control. Our result deviates from initial expectations, as electronic dance music and rock music were primarily selected to serve as aversive stimuli. The assumption was based on a lack of therapeutic effect of rock music in previous studies (Cheng et al., 2024) and a generally agitated character of these music genres. The Sonata K.448 by Mozart, in contrast, had been reported to show positive effects in numerous studies under behavioral, morphological, and physiological aspects (Kühlmann et al., 2018). A possible explanation for our result lies in the frequency properties of the music pieces. Given that mice have a hearing range in frequencies between 2 to 100 kHz, the frequency distribution of Sonata K.448, which is dominated by lower to mid-range frequencies by human standards, and an average around 820 Hz, may lie outside their optimal auditory sensitivity, potentially rendering it unappealing or even stressful, though speculative due to the absence of surveillance of physiological stress markers. The EDM mix, in contrast, has a wider spectrum of frequencies, with peaks exceeding 15 kHz, and an average of around 3,200 Hz, so mice are arguably able to perceive broader parts of the music than is possible in the Mozart composition.

Furthermore, as we have shown, the rhythm of EDM generally consists of repetitive and rhythmic beats with a clear periodicity. The music by Mozart, in contrast, is highly structured, complex, and harmonically rich. Previous work has demonstrated that, e.g., female mice are attracted to rhythmic and temporally regular acoustic courtship behavior (Perrodin et al., 2023). This may reflect an evolutionary adaptation to the environment in which complex sensory input is filtered to prioritize information relevant for survival. Accordingly, the rhythmic regularity and temporal clarity of EDM may be easier for mice to perceive, process, and interpret, potentially explaining the observed behavioral response, suggesting a relative preference of EDM over Mozart. Conversely, the relative avoidance of Sonata K.448 over the other presented stimuli might stem not from aversion per se, but from the greater perceptual challenge posed by its complexity, suggesting that music with less temporal predictability may be less accessible or less behaviorally engaging for mice under these conditions.

Notably, the initial preference findings were not further supported in the frequency test. No significant differences in the time spent in either the music or the silent room were observed, which could indicate indifference rather than a true preference. Alternatively, this may also suggest that the potentially less attractive character of Sonata K.448, or the comparatively engaging character of EDM is not frequency-dependent. Furthermore, the animals did not spend more time in a music room than in the silent room, indicating that, under conditions of short-term exposure in a novel environment, music may not be more appealing than baseline silence. However, due to the brief duration of the experiment, it remains unclear whether this lack of preference reflects a general aversion to music or is simply a result of the specific experimental paradigm. Additional behavioral measures such as rearing, freezing, locomotor speed, transitions, grooming, and defecation could provide valuable context for interpreting these findings and help to distinguish true indifference from reduced occupancy to a particular music room. As the custom setup and tracking system only extracted positional data and did not store full-video recordings, these parameters could not be analyzed retrospectively. Further research should therefore implement these complementary measures.

Nevertheless, in a direct comparison under equal conditions, mice spent significantly less time in the chamber playing Sonata K.448 compared to the other presented music stimuli, indicating a robust behavioral response. Music is perceived among humans variably, however, the variance in this result is remarkably low, highlighting the validity of the finding.

The use of a self-selection setup in our study, which allowed voluntary decision-making for a respective music genre, in combination with a habituation period prior to the measurement, likely minimized the stress-related bias.

It is important to point out that our results do not necessarily contradict previous findings on the positive influence of classical music on rodents. Numerous published studies are using rats in their experiments, which have a comparably lower frequency hearing range than mice. Additionally, recent research suggests that the analgesic effect of auditory stimulation in mice is rather based on a low signal-to-noise ratio of 5 dB than on harmonics (Zhou et al., 2022). It may be argued that auditory enrichment on a low scale can be sufficient to unfold its effect. Thus, it might not be necessary that the experimental animals hear the whole composition. Lastly, music preference should not be equated with therapeutic efficacy.

As this study is intended as a pilot study rather than a source of definitive conclusions, several limitations should be noted.

One limitation of our work is the short-term observation of animal behavior over an hour, whereas most of the therapeutic studies examine the effect of music intervention over a longer period. This constraint reflects the pilot nature of our study. Further research is necessary to examine how music preference develops over time and whether habituation or sensitization occurs with repeated exposure. Chronic and randomized exposure protocols would help to clarify true preference while controlling for novelty effects.

Another limitation is that, although minimized, cross-talk between rooms could not be completely eliminated; therefore, a potential influence of blended frequencies on our interpretations cannot be excluded.

To reduce the number of experimental animals, the same individuals participated in all three experiments. Resting periods of at least 48 h between sessions in the frequency experiment and a one-week interval between the music preference test and the subsequent frequency experiment were implemented to minimize adaptation and carry-over effects. In addition, chamber positions were randomized so that no mouse encountered the same musical conditions in the same compartment across sessions. Despite these precautions, we cannot fully exclude that animal reuse or order of experiments affected outcomes of the frequency experiments. Repeated exposure to the same auditory environment may have triggered familiarity or expectation, potentially biasing decision making by reducing novelty-driven engagement. Moreover, cumulative fatigue from multiple testing or exposure to potentially stressful musical conditions, could have contributed variability and should be considered when interpreting the results. However, the main findings from the preference test are unlikely to be influenced by these factors, as it was the first experiment conducted.

Finally, we based our study on a behavior paradigm. Further correlation with physiological markers, such as blood pressure, heart rate, or stress hormone levels, would provide a more comprehensive understanding of auditory preference in mice.

The presented assay provides a foundation for future research on auditory perception of music and its neurobiological influence in mice. Further experiments should include measurements of biological correlates of stress and well-being during music exposure and their correlation with behavior, as well as the evaluation of various ex vivo biological endpoints, with both acute and chronic exposure protocols considered. Moreover, explicitly linking behavior and neurobiological measures outcomes to specific acoustic features (e.g., rhythm, tempo tonal variability) would deepen our understanding of which musical components are most relevant for neurobiological effects. AI generated music with predefined rhythmic and spectral properties, potentially resembling the ultrasonic vocalizations of mice, could help to further explore the mechanisms underlying relative music preference. It is important to note that the music used in this paper is not fully representative of the genre labels, and that potential effects of music are likely driven by acoustic features rather than by genre itself; this should be considered in future research.

Overall, our study highlights the importance of critically selecting music used for therapeutic approaches in rodents. The common assumption that classical music is universally beneficial may not apply across all species and strains. Complex harmonics composed for human auditory perception are used in a substantial number of studies under the implicit premise that mice will perceive and derive benefit from them in a manner that is similar to humans. However, our findings indicate that, within our experimental setup, music by Mozart, specifically Sonata K.448, was comparatively less engaging than the other presented music stimuli, as measured by time spent in the compartments, suggesting that Sonata K.448 may not be the most suitable for murine therapeutic approaches.

Hence, the auditory environment should always be tailored to the respective laboratory animal, as effects may be diminished by an inappropriate choice of music. This is also emphasized by Akiyama and Sutoo (2011) reporting that the lowering blood pressure effect of classical music by Mozart is dependent on the audibility of the music, as the effect vanishes when low-pass filters are applied to the music. Van der Valk Bouman et al. (2024) demonstrated that music from a preferred genre attenuates pain in humans, regardless of the specific genre, underscoring the importance of tailoring music to the individual preference of the experimental subject.

In conclusion, our pioneer study provides novel insights into the music preference of wildtype C57BL/6J mice, suggesting that Sonata for Two Pianos in D major, K.448 by Mozart may be less engaging than selected compositions of EDM, rock or pop music. This finding challenges the assumption that music by Mozart is universally suitable for therapeutic interventions in mice, underscoring the importance of tailoring auditory stimuli to the respective experimental species. The unexpectedly higher occupancy in the EDM compartment relative to the Mozart compartment may be explained by the comparatively broader frequency spectrum and the simple, regular rhythmic pattern characteristics of EDM music. Additionally, we present a novel experimental assay for analyzing the music preferences in rodents, which can be used to evaluate different music types prior to future music intervention studies.

We acknowledge the limitations of our study and emphasize that a definitive conclusion regarding the music preferences of mice would require further analysis. Our work merely provides a platform for the scientific community to investigate this topic and suggests that Mozart may not be the preferred choice for mice.

This platform could also be valuable for exploring the therapeutic potential of music further which can serve as an effective and affordable tool without unwanted side effects, particularly as complementary support to established medical treatments. Numerous studies have reported positive effects of music in humans and rodents. However, we hypothesize that the beneficial effects may depend on music preference and, consequently, on music type.

Based on our findings, we advocate for a more nuanced and well-considered music choice in future therapeutic and behavioral studies involving laboratory animals, taking into account species-specific auditory preferences and sensitivities.