The use of biomass fuels (such as wood, cow dung, and crop residues) for household energy is quite prevalent across the developing world. Unfortunately, these sources of fuel are associated with a high degree of indoor air pollution which, in turn, has been linked to a variety of diseases. Indeed, the incidence of acute lower respiratory tract infections, which are the leading cause of death among children aged 0-5, is often associated with household fuel choices and indoor pollution. To reduce negative health and environmental effects, some policymakers have suggested the use of fuel subsidies in order to encourage households to switch from traditional biomass to cleaner fuels such as kerosene, electricity, and gas. This policy suggestion assumes that households are very sensitive to fuel prices and availability. The subject of this paper is to study the salience of fuel shadow prices (i.e., monetary and opportunity costs) on household cooking fuel decisions and, consequently, child respiratory health. Guatemala provides an interesting environment in which to study fuel costs given the wide variety of fuel choices observed across households and significant regional variation in commodity prices and penetrance of fuel markets for clean fuels (a proxy for local availability). We begin by estimating reduced form relationships between community fuel prices, fuel markets and forest cover (to proxy the availability of firewood) and household fuel choice. We also estimate reduced form models for respiratory health among children aged 0-5. Identification is a key issue here: it is possible that fuel prices or fuel market presence is correlated with unobserved community level characteristics that jointly determine child respiratory systems or cooking fuel choice. We address this issue in three ways. First, we compare the distribution of observable community and household level characteristics in areas where fuel markets for clean fuels are available to those where these are not present. We conduct a similar comparison for fuel prices. Second, we estimate reduced form models for child intestinal symptoms, receipt of prenatal care and other health symptoms and behaviors. A significant relationship between fuel prices and these other outcomes conditional on other relevant and observable community characteristics would speak towards endogeneity. Third, we exploit the seemingly random variation in the time of household interview and construct spot fuel market prices for that week/month by interacting the department of residence with world real crude oil price. We then use this new data series in estimating the reduced form models. Finally, we use the fuel price variables as instruments to look at the effect of predicted fuel choice on the incidence of respiratory symptoms, height-for-age, and weight-for-height among children aged 0-5. This is done in anticipation of on-going randomized experiments seeking to quantify the causal effect of household cooking fuels on respiratory health. The preliminary results indicate that fuel prices and availability are strongly associated with household fuel choice and child respiratory health. The identification and robustness checks reveal that this relationship likely causal: the fuel cost variables are not associated with other health conditions and behaviors or other community level characteristics. Additionally, using world fuel prices interacted with department of residence, an identification strategy which relies on plausibly random time of survey, yields similar results.