Abstract. Several large cities in developing countries have implemented driving restrictions based on number plates as a way to ration road space. Previous studies on the effects of these policies on pollution and congestion have produced mixed results. However, very little is known about how drivers directly affected by the policy change their behavior, in terms of compliance, labor supply and substitution to other transport modes. This paper uses panel data on driver behavior collected during a recently implemented driving restriction trial policy in Delhi, India. The policy restricts cars with odd license plates on even days, and vice-versa. This (effectively) randomly divides drivers into two groups, which enables a rigorous analysis of driver behavior on restricted versus unrestricted days. We find partial compliance, with a significant share of respondents reporting driving despite restrictions. We observe significant shifts to other household vehicles and public transportation options, except buses, and a decrease in trips, including work trips. Restricted days witness a small decline in trip satisfaction and an increase in trip-related inconveniences. These results show that drivers seem for the most part able to significantly alter their transportation behavior over short periods of time and with little advance notice, which leads to moderate levels of disruption.