This paper offers an affectedness-based analysis of the Chinese excessive resultative construction, which typically describes events of affectedness consisting of two participants, a theme participant and a scale participant measuring the degree of affectedness. In such an event, the theme participant is created or affected according to a beforehand prescribed value (e1) on a scale, while the process of the event results in an actual value (e2) on the same scale. The realized value may or may not be identical to the prescribed value. When the two values do not coincide (e2>e1), the 'more than expected' excessive resultative interpretation arises. This analysis crucially hinges upon the assumption that there is a covert comparison between two values on the same scale. If such a comparison cannot be established within a resultative construction, the excessive meaning will not arise.