Temporary roadworks increasingly disrupt commuting in the Netherlands, yet behavioural adaptation is still weakly represented in roadworks appraisal. This paper examines how disruption severity reshapes commuter responses, with particular attention to continued car use, public transport (PT) switching, and working from home (WFH). The analysis is based on an online survey among working commuters and combines an attribute-based stated-choice experiment, a commute-timepivoted stated-choice experiment, and a revealed-preference module on experienced major disruption (N = 180). The results show that adaptation to roadworks severity is non-linear: modest increases in car travel time are often tolerated, whereas stronger disruption triggers substantially more substitution away from routine car commuting. WFH emerges as the dominant substitute when feasible, while PT plays a secondary but meaningful role. Work-related constraints, especially WFH feasibility and on-site obligations, strongly condition the adaptation channel. Revealed responses indicate stronger car persistence than stated scenarios suggest, pointing to an intention–behaviour gap. The findings imply that roadworks demand reduction should not be treated as a uniform residual effect, but as the outcome of multiple behavioural response channels. Index Terms—temporary roadworks, disruption severity, commuter adaptation, mode choice, working from home, behavioural heterogeneity, stated-choice experiment