Although texting (or, more broadly, visual-manual interaction with a mobile phone) while driving is widely known to compromise road safety, findings from observational studies as well as surveys indicate that this form of driver distraction is more prevalent than ever, and will not disappear anytime soon. Naturally, this results in texting related crashes, which, as undesirable as they are, also constitute an opportunity for behavioural change on behalf of the driver. As the psychology of learning would suggest, experiencing negative consequences of a certain behaviour can result in a decrease of this behaviour in the future. Aim of the investigation presented in this article was to test this assumption with the help of naturalistic driving data. Thirty-nine texting related crashes (all non-injury) from the SHRP 2 dataset were analysed. Drivers’ texting behaviour in the 30 min of driving time before and after the crash was annotated from the corresponding video material, and measures such as the proportion of driving time engaged in visual-manual interaction with the phone or the driving time until the first interaction after the crash were calculated. The results provide no clear evidence for behavioural adaptation, as there was no significant reduction in the proportion of driving time spent interacting with the phone. Thirty seven of the 39 drivers involved texted again within 30 min of driving time after the crash, with a median of only about 2 min. The findings indicate that the experience of (comparatively minor) crashes might not be sufficient to deter drivers from texting while driving.