Longitudinal effects of intergroup contact on prejudice were investigated in a sample of 65 young adults (Sample 1) and a sample of their close friends (Sample 2, N= 172), adopting a full cross-lagged panel design. We first validated the self-report measure of intergroup contact from Sample 1 with observer ratings from Sample 2 by demonstrating that self-reports and observer ratings of contact were highly correlated. Moreover, we obtained significant cross-lagged effects of intergroup contact on prejudice with both contact measures, thereby providing a second validation for the use of self-reports of intergroup contact. Finally, by the use of latent change modelling, we demonstrated that, although no overall significant change in contact and prejudice over time was found, there was meaningful variation in absolute change in the individual levels of intergroup contact and prejudice. In particular, some individuals showed increases while others showed decreases in contact or prejudice across time. Moreover, higher levels of intergroup contact at Time 1 were followed by larger subsequent decreases in prejudice between Time 1 and Time 2, and changes in contact were significantly and negatively related to changes in prejudice. Methodological implications of the findings are discussed.