Two laboratory studies with 332 student participants investigated secondary confessions (provided by an informant instead of the suspect). Participants allegedly caused or witnessed a simulated computer crash, then were asked to give primary or secondary confessions during interrogation. Study 1 replicated the false evidence effect for primary confessions. Secondary confessions were obtained at a high rate, which was increased by false evidence in combination with incentive to confess. In Study 2 a confederate either confessed to or denied crashing the computer. Incentive increased the rate of secondary confession only in the presence of a denial; that is, incentive increased the number of false secondary confessions only. Implications for the use of incentives during informant interrogation are discussed.