For an environmental review to predict project impacts to wildlife, the wildlife community must be accurately characterized as part of the existing environmental setting. Biologists seek to accomplish this by reviewing occurrence records and with various types of survey – the most common being reconnaissance survey. However, reconnaissance survey objectives vary, as do methods, interpretation of results, and reporting. We propose a protocol for performing wildlife reconnaissance surveys to maintain flexibility in survey conduct while standardizing specific field and reporting methods to enable the establishment of benchmarks against which to compare and interpret survey findings, and to determine whether the wildlife community has been sufficiently sampled in support of an accurate characterization. We propose that each new species detection be recorded with time into the survey, and that the number of species detected be modeled as a function of time into the survey. The model can predict the number of species theoretically available to be detected during the survey, and it can predict the number of species detected at standardized survey durations. We offer recommendations on objectives, how many surveys to perform, survey duration, times of day to survey, personnel, field techniques, documentation of detections, minimum reporting levels, and interpretation of results.