Standard cost-benefit analyses of transport policy measures will not capture all benefits and losses if there are labour market imperfections. In the case of congestion charges, theoretical analyses have raised concerns that these effects may constitute considerable losses, possibly to the extent that aggregate welfare is reduced, contrary to conventional wisdom. We investigate this by estimating the effects on labour income of the Stockholm congestion charges, using an estimated relationship between accessibility and income. Results show that effects on labour income are, in fact, positive. It turns out to be crucial that the model accounts for value-of-time heterogeneity.