The Dutch turbo roundabout is characterized by predictable lane use and radial connecting entries. To calculate the capacity, in 1997 the multilane roundabout explorer was developed by modification of the capacity model of Bovy, which considered both separated lanes and pseudoconflict. Because of its linear structure, this model did not take into account multilane roundabout properties in a good way. Therefore a new model was developed by modification of the model of Hagring. Using only the gap acceptance approach to calibrate the parameters is not sufficient because the pseudoconflict is not taken into account. A combination with a stream approach is necessary. For validation of an analytic model, it is important to realize that the distribution of gaps offered in two flows simultaneously is determined primarily by a stochastic process. Thus, it is possible to test the effect of the distribution of the flows over two circulatory lanes by using a microsimulation model. The parameters needed to tailor VISSIM microsimulation software for this purpose were estimated. It is concluded that a VISSIM model may be suitable for the analysis of both single-lane and two-lane roundabouts. This paper deals with this combined approach. The findings are that (a) the pseudoconflict should be taken into account and (b) the turbo roundabout will have a larger capacity than the Dutch standard concentric double-lane roundabout because of the better circulatory lane-split.