To predict maximal exercise capacity, we have developed the rating of perceived capacity (RPC) scale, based on metabolic equivalents (METs). MET values from 1 to 20 were listed on a progressive scale and linked to physical activities. Test subjects rated their perceived capacity by choosing the most strenuous activity and the corresponding MET value that they could sustain for 30 min. In order to validate the RPC, 87 healthy women participated in the study (age range 21-79 years). The capacity rated from the scale (MET(RPC)) was compared with the actual work capacity in a maximal incremental ramp cycle test. The maximal load achieved at the cycle test (in W), was transformed into the unit MET (MET(test)). The mean difference between MET(RPC) and MET(test) was -1.4 (3.18) (2SD). Multiple linear regression showed that MET(test) depended significantly on MET(RPC) and age. Extended regression analysis suggested that a nonlinear correction was most appropriate in order to improve the prediction. The maximal predicted capacity was given by: MET(pred) = (5.08 + 0.70 x MET(RPC))/[1+(e(0.059 (age-87.2)))]. After this correction, the mean difference between MET(pred) and MET(test) was reduced to 0.0 (2.62) (2SD). Reference values for MET(pred) and MET(test) were also obtained. In conclusion, the RPC combined with an age correction was shown to predict accurately maximal exercise capacity; it is also quick and easy to use. Thus, RPC may be a valuable tool in the prescription of exercise and in monitoring the effects of exercise and rehabilitation.