The effect of a road improvement or new road in a city can be analysed as follows. Thefirst-round effects are that the new or improved road will cause a redistribution of traffic in the city's network of roads. Some roads, mainly those leading to or from the new road, will become more congested-i.e., speeds will fall and vehicle costs, in terms of time and operating costs, rise. On other roads congestion will be diminished because on balance traffic will fall owing to the diversion of traffic to take advantage of the new or improved road. The secondround effects are that since the network capacity is increased, traffic will be attracted, both to the improvement itself and to decongested roads. Traffic will save time and have lower operating costs than it would have had if the improvement had not been made. This may be called cost-fall-induced traffic generation. If the road improvement is small or the benefit is diffused over a large area, experience suggests that cost-fall-induced traffic generation will be proportionally less than if the improvement is large and noticeable. In the extreme case of an urban expressway system, generated traffic may grow to be 30 per cent. or more of the traffic using the new or improved facility; and where there is decongestion elsewhere in the network, the like reduction in the price of travel will also induce new traffic. And there will also be trend growth in traffic, i.e., traffic which would have appeared whether or not the improvement was made: because the demand for the services provided by the network may grow through time. Calculations of rates of return on road investment-or of costbenefit ratios, etc.-have run in terms of time -and vehicle cost savings to traffic. Many of these have not adequately taken into account the first-round effects.' None we have noticed estimates the magnitude of