The authors consider the problem of scheduling jobs on parallel related machines owned by selfish agents. They present a family of algorithms that yield a (4+e)-approximation that incentivizes each selfish agent to report the actual speed of his or her machine -- i.e., each agent will maximize his or her own profit by reporting their actual speed.