The safety implication of new in-vehicle technologies is a leading concern for car manufacturers. Several methods aim to measure the driver distraction induced by driver information and assistance systems. One of these methods, denoted the Lane Change Test (LCT), aims to measure quantitatively the degradation of the driving performance induced by secondary tasks. An experiment involving 17 participants was conducted from September to November 2006 to investigate the robustness of the method. A calibration task was used to compare performances in PC and in simulator environments. Radio and navigation tasks were performed in four different vehicles to assess the relevance of the method to discriminate among different types and location of in-vehicles devices and displays. In addition to the main indicator suggested in the LCT procedure (mean lateral deviation), features of the secondary tasks (latency, duration) were considered. The results confirm the transferability of the method from PC to vehicle-based environment, but question the sensitivity of its main indicator to discriminate between vehicles and functions