This paper analyzes one of the uncertainties defined for the Dutch long term strategic policy study “National market and capacity analysis.” The question asked was whether the upcoming of self-driving cars would impact the investment needs in the trunk road system for the years 2040 and beyond. The paper describes how the National Model System was adapted to do this analysis, how the scenarios were defined, and which results were obtained. The main lesson from this study is that a “naïve” exploration of this question, only assessing the first order effects on congestion, results in a large over-estimation of the reduction (or increase) in congestion. It is crucial to model mode, time of day and destination effects as well, since that will significantly reduce the first order effects on congestion.@en