Participants were randomly selected from arrest records in jails and booking facilities in major metropolitan areas; no personally identifying information is taken from each record chosen, and prevalence estimates can be generated regarding those arrested in the catchment area covered by the jail.
ADAM began as the Drug Use Forecasting program in 1987, which tested arrestees in 13 (later 23) jurisdictions on a quarterly basis.
Some of the smaller locations such as Woodbury County, Iowa and Rio Arriba, New Mexico were added late in the program's life, in an attempt to gain information about non-metropolitan areas.
In 2007 ten of the original sites continued: NYC, Chicago, Washington DC, Denver, Charlotte, Indianapolis, Sacramento, Portland, Minneapolis, Atlanta.
In the last two years of operation, the program continued in five of these sites: Denver, Sacramento, NYC, Atlanta, Chicago.