I have a question regarding the terminology used when specifying power factor for sources vs. loads. For inductive loads, the power factor is assumed to be *lagging*, with positive Q, and "absorbing" VARs. For capacitive loads, the power factor is assumed to be *leading*, with negative Q, and "producing" VARs. This much makes sense. What I don't understand (or haven't been able to verify) is the terminology regarding the source - let's say a generator - that is powering these loads. If there is a net lagging load, then the generator is set to be overexcited, i.e. *lagging* or "pushing VARs", and vice versa. This seems contradictory to me. Given that a *leading* capacitive load will balance out a *lagging* inductive load for power factor correction, why is it that a *lagging* generator will balance out a *lagging* inductive load? Is the terminology just reversed when discussing the source vs. the load? Additional question: assuming my statement above is true, does the current exiting the generator lead or lag the voltage for correcting an inductive load?