This paper comments on the recently published guidelines of the Washington Panel on incorporation of indirect non-medical costs, or productivity costs, in economic evaluations of health care. Traditionally the human capital or more recently the friction cost method is used to measure these costs. The Panel, however, recommends incorporating productivity costs as health effects in the denominator of the C/E ratio. This paper argues that incorporation of productivity costs in cost-effectiveness analysis expressed as health effects is not correct. Only direct health related effects on quality of life that cannot be meaningfully monetarized should be considered as health effects. Furthermore, measuring productivity costs in terms of quality of life may lead to misrepresentation of these costs from a societal viewpoint. This misrepresentation occurs because of the existence of social security systems and private insurance compensating for income reductions from disease. Furthermore, the patient's viewpoint is useful for quality of life measurement, but not for measuring productivity costs from a societal perspective. Finally, alternative recommendations are formulated for incorporating societal productivity costs in economic evaluations of health care.