In an experiment on sequential rent seeking, we find that the theoretically predicted first-mover advantage does not exist empirically. Rather, a second-mover advantage can be observed. Though the subgame perfect equilibrium can not prevail in experiment, individual behavior is highly reasonable nevertheless, such that subjects appear to play a reduced form of the original game. We also observe that the efficient fair outcome can not be achieved because of the exploitation opportunity for the second mover. The observed behavior is quite different from that in ultimatum and trust game experiments.