In developmental and reproductive toxicity studies, drinking water is a common means of delivering the test agent. Reduced consumption of toxicant-containing water raises questions about indirect effects of reduced maternal fluid consumption resulting from unpalatability, versus direct effects of the test compound. Issues to consider include: objective assessment of dehydration and thirst, the relative contributions of innate and learned behaviors to drinking behavior and flavor preference, and the objective assessment of physiologic stress. Not only do lab animals under ad lib conditions consume more water than the minimum required to maintain fluid balance, animals faced with water restriction have substantial physiologic capacity for protection of metabolic processes. Measures of blood biochemistry can provide quantifiable, objective indications of fluid balance, but changes in these parameters could result from other causes such as effects of a test toxicant. Consummatory behaviors in response to perceived need are highly influenced by learning. Hence, the drinking behavior, water intake, and flavor acceptance/preference of animals used in toxicology experiments could be subject to learning experiences with the test compound. Physiological symptoms of stress produced by water deprivation may be distinguishable from the symptoms associated with other generalized stressors, such as food deprivation, but doing so may be beyond the scope of most developmental or reproductive toxicity studies. Use of concurrent controls, paired to test groups for water consumption, could help distinguish between the direct effects of a test toxicant as opposed to effects of reduced water consumption alone.