We read with interest a recent report by D'Andrea and Reddy describing differences in hematologic and hepatic blood profiles among a cohort of workers involved in Gulf oil spill cleanup activities relative to an unexposed cohort. They report that cleanup workers had lower average levels of platelets, blood urea nitrogen, and creatinine, and higher levels of hemoglobin, hematocrit, alkaline phosphatase, aspartate aminotransferase, and alanine aminotransferase. The authors also postulate that "the oil spill exposure appears to play a role in the development of hematologic and hepatic toxicity." The relation of these findings to exposures associated with cleanup activities cannot be interpreted given the current study design, lack of exposure history, and laboratory analysis. With regard to the study design, we have concerns about both the validity and generalizability of the results. Blood chemistries from workers participating in cleanup activities along the coast of Louisiana for 3 months or longer are compared to a cohort of randomly selected patients visiting a clinic for a wellness checkup, located 100 miles away from the Gulf coast of Louisiana. Additional selection and exclusion criteria, such as employment history, demographic and lifestyle characteristics, and medical history are not included. This recruitment design provides little assurance that the 2 groups come from the same base population or that cleanup workers chosen for this study are representative of the broader group of cleanup workers. The use of a well-characterized control group, such as workers with similar demographic, lifestyle, and medical characteristics, without exposure to cleanup activities, or an appropriately matched external referent group of subjects selected from the National Health and Nutrition Examination Survey cohort would have been preferable. The current study design also does not incorporate matching on demographic characteristics, and as the authors acknowledge, the exposed and unexposed groups differ markedly on 2 key covariates-age and sex. In their analysis, the authors attempt to address each of these factors separately, but fail to account for the effects of age and sex simultaneously. The stated limitations also are incomplete with regard to consideration of additional confounders. Hematologic and hepatic biomarkers levels vary based on demographic characteristics and are affected by lifestyle exposures. Additional covariates, such as race, ethnicity, chronic medical conditions, medication use, smoking, and alcohol use are not reported or analyzed and may represent important confounders. Information regarding the hazard, source, and potential for exposure is incomplete. The duties performed by the cleanup workers, or use of personal protective equipment, are not specified. This information is important because offshore activities (eg, in situ oil burning) and onshore activities (eg, wildlife cleanup) differed greatly in their exposure potentials.5 Similarly, routine use of safety glasses, gloves, sleeve protectors, rubber boots, aprons, and other protective clothing would be expected to mitigate exposures. Additional exposure information, such as pre- and post-cleanup work exposures and occupational history, is necessary to understand the relationship between cleanup activities and potential differences in laboratory values. Adequate characterization of exposures is critical; postdisaster research without adequate exposure characterization is unlikely to result in clear conclusions about health outcomes. With respect to biological specimen collection, no information is provided about the procedures used in the collection of specimens, timing of collection, storage of specimens, and how specimens were analyzed. How the specimen was collected and which type of blood collection tube was used can affect results. Timing of the collection after exposure is important when measuring urinary biomarkers because the optimal collection time is dependent on the kinetics of the biomarker. When considering the standard deviation, many of the mean values for the hepatic biomarkers reported are within the normal range for these assays, making interpretation of any differences difficult without additional information concerning the exposure of interest, other exposures that may have been present in the environment or diet, and lifestyle factors. Understanding the health effects of participating in disaster mitigation is important for workers, their families, and the affected communities. Greater attention to study design, exposure history, and laboratory analysis are necessary before concluding that differences in blood profiles between workers involved in Gulf oil spill cleanup and another cohort of interest can be attributed to cleanup work.