Recent literature has shown that lying behavior in the laboratory can well be explained by a combination of lying costs and reputation concerns. We extend the literature on lying behavior to strategic interactions. As reciprocal behavior is important in many interactions, we study a theoretical model on reciprocity where a player's altruism depends on her perception of the other player’s altruism towards herself. We analyze a sequential two-player contest and vary the second mover’s information on the first movers lying behavior. This allows us to derive predictions on the second mover’s behavior which we test empirically in a large scale online experiment and in the laboratory. In both experiments, the second mover’s lying propensity does not depend on whether the first mover has (possibly) lied or not. This robust behavioral pattern provides strong evidence that reciprocity does not play a role for lying behavior in our setting.