Biological rhythms govern a wide range of behaviours and physiological processes. In behavioural neuroscience, patterns such as locomotor activity or sleep-wake cycles are analysed to understand endogenous timekeeping mechanisms. In circadian rhythms, the timing of behavioural events relative to internal or external cycles is key to understanding how clocks respond to environmental cues. Under constant conditions, these rhythms most clearly reveal the intrinsic properties of the clock. Traditional analyses often focus on immediate responses to external cues, which can obscure long-term phase stability and complicate comparisons across groups, such as different genotypes or experimental conditions. To address this, we developed an analytical pipeline for assessing the behavioural phase and estimating phase lags under prolonged constant conditions. The method extracts oscillatory components directly from behavioural data without referencing external cues or imposing rigid period models. Implemented in R, with an open and extensible framework, the pipeline includes tools for data formatting, signal filtering, visualisation, statistical comparison, and guidelines for evaluating preprocessing choices. Although illustrated using Drosophila melanogaster locomotor activity, this approach applies to any oscillatory biological variable, enabling robust phase comparisons and providing a versatile and reproducible tool for studying oscillations.