Repeated one-shot public good experiments commonly tell participants only oftheir group's total contribution after each round. In contrast, private charities sometimes publicise large contributions or contributors to encourage others to give or to bring recognition to donors. The effect of supplying such selective information on contribution levels is tested here experimentally. Following a control treatment with standard information, a second treatment also informs subjects of the maximum contribution made in their group after each round. In a third treatment, subjects are further given the opportunity to make costly rewards to the (unidentified) maximum contributor. Revealing generous contributions appears to raise average contributions slightly. Adding the ability to reward large contributors does not generate further increases, but raises the variance of contributions.