The Stockholm toll causes, as predicted by theory, a reduction in traffic, leading to increased speeds, and to time gains for remaining car-users. These gains, calculated to be about +110 M. SEK per year, appear to be modest, much lower than similar gains estimated in London, because congestion was moderate and reducing it to its optimal level, which is what the toll achieves, does not represent large increases in speed. The toll also causes environmental benefits, for an estimated +102 M SEK per year. On the other hand, the toll causes a loss for evicted car-users, for about -66 M. SEK per year. A major cost is the implementation cost, about half the cost experienced in London, but nevertheless high at about -730 M SEK per year. Finally, the toll made it necessary in order to accommodate modal shifters to increase public transport supply, at a cost of - 580 M SEK per year, although this increase in public transport supply was not sufficient to prevent a deterioration in service quality tentatively estimated to be above 200 M SEK per year. Overall, costs outweigh the very real benefits of the toll by more than one billion SEK per year. For an urban toll to produce net benefits, it seems that three conditions are required: a relatively high degree of congestion, a reasonably cheap implementation system, and a public transport system with a low marginal cost.