In this paper, we explore the deep economic crisis experienced by the Turkish economy in 2001 and 2008 as quasi-experiments to causally identify the association between economically challenging conditions in-utero and child's birth weight. First, we utilize the temporal and spatial variations of the economic crisis. Second, we estimate mother fixed effects models by exploring the variation in the birth weight of siblings by their in-utero exposure to economic downturn. Using the Turkish Demographic Health Surveys (DHS), we find that a higher regional GDP contributes significantly to better birth outcomes among crisis-time children; adverse health effects are mainly observed in children born to mothers with low socio-economic status, suggesting that the main driver of the estimated effects of economic crises in Turkey is credit constraints. Our mother fixed effects models reveal that selective fertility, abortion, and unobserved heterogeneity across mothers are important omitted variables in the interpretation of regional regressions. The estimated effects of economic downturns cease to be statistically and economically significant once we control more accurately and directly for a family's firsthand experience with economic recessions. Thus, our results demonstrate that regional-level regressions estimating potential infant health costs of economic recessions potentially overestimate the true effects, and more direct measures of unobserved heterogeneity should also be considered in these analyses.