Dishonest activities with negative consequences for others and society are often undertaken by individuals as well as groups of people. In this paper, we use a field experiment among students aged 11-16 years to study whether there is a difference between individual and group cheating behavior. We find that students cheat, but not to the maximum extent possible. On average, groups are more inclined to cheat than individuals, but there are important differences across age. While there is no evidence of dishonesty among younger individuals, older individuals as well as younger and older groups cheat and do so to a similar extent. The way in which groups are formed does not seem to matter.