How are experiences of perceived exertion (PE), associated with varying load, integrated in a single measurement? In search of an integrating pattern of scalings of PE, 209 participants were randomly assigned to 7.5-min pedalling trials on a bicycle-ergometer. Intensities were either kept constant at 25W, 50W, 75W, 100W, 125W, or were systematically varied after 2.5 and 5min whereby the overall load was kept constant at either 50W, 75W or 100W. Systematically varied intensities were either continuously increased or decreased by steps of 25W. A nearly linear relationship between steady presented intensities and session scalings of PE confirmed the validity of the Category partitioning (CP) procedure. Scalings obtained in sessions with systematically varied loads were altered by the order of the intensity changes. The influence was more pronounced in sessions with increasing loads than in sessions with loads in decreasing order.