The concept of detection limits dates back to the 1960s, with the best-known treatise being Lloyd Currie’s 1968 definition of “detection limit” in terms of a “critical level.” The concept is important for scientists, policymakers, instrument manufacturers, and other stakeholders. It remains a source of controversy, however, because although the theory is based in statistics, it is applied by analytical practitioners. This is not always compatible. Over the past half-century, there have been many formulations, some developed for particular programs, and with these formulations, a cornucopia of terms and abbreviations, presenting challenges for a laboratory dealing with the requirements of different programs. Currie’s principles were solidified in the 1995 International Union of Pure and Applied Chemistry recommendations and International Organization for Standardization 11843 series. These principles have been embraced in Europe, but not in the United States. One possible reason is that the concepts focus on underlying statistics and not on practical application. The goal of ASTM Committee D22 on Air Quality is to begin with the smallest common denominator, a well-behaved air quality method and then to standardize a practice that has statistical rigor; will be understood and used by the analytical community, and, perhaps most important, will produce consistent results. Committee D22 hopes to expand on this admittedly narrow focus in the future. This paper describes fundamental concepts of the Committee D22 methodology, based on Currie’s 1968 and 1999 papers. It uses a one-sided test of significance to affirm or reject the null hypothesis (analyte not detected). Rejection of the null is used to determine a value above the mean background that represents detection of the analyte signal with a stated probability of repeatability. It evaluates all significant error sources and uses the magnitude and variance of the background and combined error to estimate the detection limit.