This paper explores the link between the prevalence of violent conflicts and extremely low female labor force participation rates (FLFPR) in South Asia. We merge Labor Force Surveys (LFSs) from Bangladesh, Sri-Lanka, India, and Pakistan to the Global Terrorism Database (GTD) to estimate the relationship between terrorist attacks and female labor supply. We exploit the availability of geographical level data on exposure to violence, comparing administrative units exposed to attacks with administrative units not exposed. We find that one additional attack reduces FLFP rates by about 0.008 percentage points, on average. Violence has less impact on male labor participation, thus widening the gender labor participation gap. Also, one extra wounded person or one extra killed person reduces FLFP rates by 0.0015 and 0.0048 percentage points on average, respectively. We test the added-worker effect theory - which posits that violence might increase FLFP as women try to make up for lost household income - and find mixed evidence: greater prevalence of attacks may encourage married women to exert more working hours, but when the environment gets more risky as number of dead and wounded people increase, all women work less hours. We also test the non- linearity of various violence effects, finding that violence decreases FLFP less where FLFP was already higher before the advent of violence, and that violence has a progressively greater impact on lowering FLFP where the number of attacks is higher.