The general aim of the thesis is to answer questions on general and differential aspects of perceived exertion and on the measurement of its intensity variation. Overall perceived exertion is commonly treated as a unidemensional construct. This thesis also explores its multidimensional character. Four empirical studies are summarized (Study I-IV). Psychophysical power functions of perceived exertion obtained with the new improved Borg CR100 (centiMax) scale were found to be consistent with results obtained with absolute magnitude estimation, and with the classical Borg CR10 and RPE scales. Women gave significantly higher perceived exertion scale values than men for the same levels of workload on a bicycle ergometer. This agrees with the fact that they were physically less strong than men. With regard to the measurement of “absolute” levels of intensity, RPE- and CR-scale values were validated by physiological measurements of heart rate and blood lactate. Predicted values of maximal individual performance obtained from psychophysical functions agreed well with actual maximal performance on the bicycle ergometer. This confirms the validity of the RPE and CR scales for measuring perceptual intensity and their value for interindividual comparisons. To study the multidimensional character of perceived exertion, 18 symptoms were measured with a CR scale: in a questionnaire, and in bicycle ergometer work tests. Five factors were extracted for the questionnaire: (1) Muscles and joints; (2) Perceived exertion; (3) Annoyance/lack of motivation; (4) Head/stomach symptoms; and (5) Cardiopulmonary symptoms. Four factors were extracted for the bicycle max test: (1) Physical distress; (2) Central perceived exertion; (3) Annoyance/lack of motivation; (4) Local perceived exertion. The questionnaire is suggested for clinical use to let patients express a variety of symptoms. The thesis also resulted in improvements of the Borg CR100 scale. An extended use of the scale is recommended.