Changes in some vaginal mucus parameters were studied in order to generate predictive models capable of enhancing oestrous cycle staging, using equal groups (unsynchronized‐USC [no treatment] and synchronized‐SC [Synchromate® i/m on d0, d11]) of Bunaji cows (n = 48) aged 3–4 years. Vaginal mucus was collected (starting d11 in SC) daily over 26 days using standard procedures. Physical (viscosity, elasticity, density, resistivity) and biochemical (pH, glucose, cholesterol, total protein, calcium, magnesium, sodium, potassium) parameters were evaluated using standard procedures. Data were analysed using chi‐square and multinomial logit regression modelling. Models generated using oestrus as reference categories were ascertained for accuracies. Chi‐square values for viscosity, elasticity and density were significant (p < .01) in USC and SC across stages of the cycle. Results for USC showed that pH and cholesterol were predictive (p < .01) for pro‐oestrus, metoestrus and dioestrus, while total protein was predictive (p < .01) for dioestrus only. Similarly, magnesium was predictive (p < .05) for pro‐oestrus. For SC, pH, magnesium and cholesterol were predictive (p < .01) for pro‐oestrus, metoestrus and dioestrus, while total protein was predictive (p < .01) for pro‐oestrus and dioestrus. Potassium and total protein were also predictive for metoestrus at 10% and 5% significance levels, respectively. Though findings suggest the usefulness of magnesium in staging the oestrous cycle only in synchronized cows, pH, total protein and cholesterol appeared to be the more important vaginal mucus parameters in Bunaji cows, regardless synchronization. Furthermore, the models developed showed high accuracy levels for staging the oestrous cycle in USC (100%) and SC (89%).