For an environmental review to predict project impacts to wildlife, the wildlife community must be accurately characterized as part of the environment. Ideally, field surveys would inventory the wildlife community, but inventory requires many surveys using multiple methods at great time and cost. Habitat assessments are performed in reconnaissance surveys and desktop reviews to determine likelihood of occurrence of species on a project site, but they vary greatly in assumptions, methods, and accuracy. We propose standards for habitat assessment, starting with a standard definition of habitat, and the assumption that habitat is present until proven otherwise. Multiple reconnaissance surveys should be conducted at various times of day (and night) across seasons. All occurrence databases should be reviewed; however, absence of occurrence records is not evidence of species’ absence. When comparing habitat associations to vegetation cover on site, refrain from pigeon-holing species into unrealistically narrow portions of the environment. Most wildlife are mobile and use more of the environment than often expected, and both the delineation of vegetation cover and the summary of habitat associations are prone to error and bias. While focus is usually on special-status species, all species compose the wildlife community, which needs accurate characterization to predict impacts.