In this study we estimated the indirect costs of back pain in 1991 in The Netherlands on the basis of two approaches: the traditionally used human capital method and the more recently developed friction cost method. The indirect costs of illness were defined as the value of production losses of paid labour and related costs to society due to back pain. The results of this study in 1991 in The Netherlands show that the short-term indirect costs estimated by the human capital method were more than three times as high as the indirect costs estimated by the friction cost method (US$ 4.6 billion vs. USS 1.5 billion, respectively). The lower estimate of indirect costs when using the friction cost method is mainly due to the fact that in this method actual production losses are estimated during a relatively short friction period, which is defined as the period needed to restore the initial production level. In contrast with the human capital method, long-term absenteeism and disability do not induce additional costs when applying the friction cost method. Since the friction cost method takes into account that employees can be replaced, we believe that this method produces a more accurate estimate of indirect costs than the human capital method. Notwithstanding the resulting decrease in indirect costs of back pain, these costs are still impressive, representing 0.28% of the GNP in The Netherlands in 1991. As a consequence, but particularly stimulated by structural changes in the Dutch social security system, policies aimed at reducing indirect costs of back pain, increasingly concentrate on the development and evaluation of interventions early after the onset of disease. This is complemented, on the one hand, by the development of clinical guidelines for the management of back pain in primary care and, on the other hand, by governmental policies aimed at reintegration of chronically ill in the labour force.