Distinguishing causation from correlation is a central challenge in empirical economics. Natural experiments generate useful exogenous variation in variables of interest that can be used to plausibly establish causation. In this thesis, I exploit natural experiments to address three distinct empirical questions of policy and research interest. Chapter 1 uses the exogenous variation in health at birth induced by drought shocks in rural India to investigate the long run effect of early life health status on height. The focus is on caste heterogeneity in the impact of drought. Caste is a strong proxy for income and wealth. I find drought at birth is associated with a 0.3 cm drop in the height of upper caste women but a 0.4 cm gain in the height of a major low caste group - the Scheduled Tribes. These findings are consistent with a model of selection which predicts that for lower castes, only the relatively robust babies survive or perhaps a more favorable socio-economic mix choose to select into fertility. I find that there is no caste gradient in height outcomes for more recent birth cohorts, suggesting that lower castes may have improved their ability to weather drought shocks. Chapter 2 investigates the link between cell phone use while driving and crash risk, an area of active research and policy concern. Most studies have concluded that cell phone usage increases crash risk. I exploit a natural experiment induced by a popular feature of cell phone plans in recent years - the discontinuity in marginal pricing at 9 pm on weekdays when plans transition from "peak" to "off-peak" pricing. There is a clear spike in call volume at 9 pm on weekdays but no evidence of a concurrent rise in crash rates. The triple difference estimator that uses weekends and the pre-cell phone era as additional controls is negative, about - 4%. The estimates are relatively precise and the upper bound rules out any positive effect of cell phone usage on crash risk. Chapter 3 investigates the impact of investor awareness on stock returns. Inclusion or deletion from the Fortune 500 list potentially changes investor awareness but is otherwise an 'information-free' event. The Fortune 500 list is a prestigious annual ranking of the 500 largest US corporations by revenue. During 1995-2004, I find evidence of a significant positive effect on returns of new entrants following the publication of the list but no effect on returns of firms that drop out of the list. The findings are consistent with a story of a one time boost to returns due to heightened investor awareness in the weeks following publication.