In order to characterize a solid surface, the commonly used approach is to measure the advancing and receding contact angles, i.e., the contact angle hysteresis. However, often an estimate of the average wettability of the solid–liquid system is required, which involves both the dry and wetted states of the surface. In this work, we measured advancing and receding contact angles on six polymer surfaces (polystyrene, poly(ethylene terephthalate), poly(methyl methacrylate), polycarbonate, unplasticized poly(vinyl chloride), and poly(tetrafluoroethylene)) with water, ethylene glycol and formamide using the sessile drop and captive bubble methods. We observed a general disagreement between these two methods in the advancing and receding contact angles values and the average contact angle determined separately by each method, although the contact angle hysteresis range mostly agreed. Surface mobility, swelling or liquid penetration might explain this behaviour. However, we found that the 'cross' averages of the advancing and receding angles coincided. This finding suggests that the cross-averaged angle might be a meaningful contact angle for polymer–liquid systems. Hence, we recommend using both the sessile drop and captive bubble methods.