To successfully get around in traffic it is often necessary to keep track of several relevant targets at the same time. This can be done by combining foveal and peripheral visual information sampling. Especially if no detailed input is needed, for example when confirming the absence of road users, it may be enough to use peripheral vision only. Using a driving simulator with an urban scenery, 35 participants passed three zebra crossings with a) no pedestrians nearby, b) pedestrians standing nearby and c) pedestrians nearby of whom one started walking towards the street. In the last case, all participants foveated the walking person, albeit around one third of the participants already released the throttle before the first glance at the pedestrian. The standing pedestrians were foveated in almost all instances, whereas the roadside nearby the zebra crossing without people nearby was not foveated by around a quarter of the participants. Taken together, the results indicate that peripheral vision may suffice to confirm the absence or presence of pedestrians. With people present, a glance towards them is initiated, likely to check for additional information. Throttle release before foveation is an indication that the walking pedestrian was detected as relevant with peripheral vision.