Advanced traveler information systems (ATIS) are computer-based in-vehicle systems that provide travel-related information to automobile drivers. It is important for user acceptance to occur quickly because ATIS performance is optimized when many automobiles are equipped with the systems. One important factor in user acceptance is how much users trust the technology. Previous studies on this topic revealed that inaccurate congestion information reduced the users’ trust in ATIS and limited compliance with ATIS advice. The previous research is extended by the evaluation of both trust in ATIS and compliance with ATIS advice in a high-fidelity driving simulator. Participants drove through a simulation with the goal of avoiding congestion. A simulated ATIS recommended which way to go to avoid congestion. Participants drove through four trials of ten intersections each. In one trial, ATIS provided accurate information at 100 percent of the intersections. In the other three, the information was 80, 60, or 40 percent accurate. The four trials were presented to the participants in one of three orders. Dependent measures included ( a) how much participants trusted ATIS advice, ( b) how many times they complied with ATIS, and ( c) how accurate they thought the system was. The results show that 40 percent accuracy would not support user trust and compliance, but that 60 percent accuracy probably would. Also, the order in which participants experienced the trials significantly affected both trust and compliance; the effects of accuracy order were similar for both trust and compliance.