Sleep–wake behaviour in mice is known to interact with various behavioural dimensions. Therefore, it is necessary to control for such dimensions when evaluating sleep in mice. The characterisation of sleep in rodents usually is based on EEG signals. Since this method demands the invasive implantation of electrodes, it cannot be integrated into general behavioural phenotyping procedures. Thus, non- or minimum-invasive methods are needed for the analysis of sleep–wake behaviour. Although physiological parameters, like for instance general locomotor activity, allow for the assessment of sleep–wake behaviour in mice, existing methods lack reliability especially in measuring stationary and three-dimensional activities. In this study, a small magnet was implanted subcutaneously near the neck muscles of mice and each movement of the magnet was registered via a sensor plate. For validation of the described method, the effects of sleep deprivation were evaluated by both the magnet and the EEG in parallel. Our results show that the data obtained via the subcutaneously implanted magnet represent a reliable and sensitive measurement of quantitative aspects of sleep–wake behaviour: spatial variation as well as stationary activities could be dissociated from sleep. Qualitative sleep characteristics were not detected. In summary, this minimum invasive method allows for the detection of quantitative alterations in sleep–wake behaviour in mice, thus, offering a useful, rapid pre-screen in animal sleep research.