This book employs quantitative methods to test the effect of human rights abuses on civil war onset. Its point of departure is the classic theories of collective violence, namely the deprived actor (DA) and rational actor (RA) research traditions. I find that in ethnically fractionalized societies, state repression decreases the risk of civil war. This gives support to the RA assumption that repression (which produce grievances) in ethnically fractionalized societies (which are the most conflict prone) should not lead to an increased probability of civil war onset. Thus, grievance is not the deciding factor for why people resort to violence.