We evaluated the utility of address- versus ZIP code-level data for determining the association between heat waves and preterm birth (PTB) or non-accidental death (NAD). Addresses on birth (1990-2010) and death (1997-2010) records from Alabama, USA during May-September were geocoded, and a time-stratified case-crossover design was implemented. ZIP code- and address-level heat wave indexes (HIs) were derived using 12.5 km air temperatures from the North American Land Data Assimilation System (NLDAS). We downscaled NLDAS data, using land surface temperatures (LST) from the Moderate Resolution Imaging Spectroradiometer (MODIS) product, to estimate finer spatial resolution (1 km) HIs in address-level analyses. Both ZIP code- and address-level analysis results show that the association between heat waves and PTB or NAD was significant and positive. These results suggest that ZIP code- and address-level analyses produce similar effect estimates with similar precision. When analyses were stratified by LST, a decreased trend on the odds of PTB and NAD during heat wave days was found from the first tertile (the lowest LST) to the third tertile (the highest LST). This result suggests that heat waves may be more problematic in lower LST areas, which are primarily non-urban.