Different methods have been developed in clinical abuse liability test
ing in man. Tolerance, psychic and/or physical dependence must be inve
stigated through clinical studies during drug development of a new sub
stance. Adequate methodology is needed using double-blind, time-blind
evaluations, comparisons of different dose levels and duration of trea
tment for a given drug, abrupt and gradual interruption of treatment,
appropriate period of observation after treatment cessation... The opt
imal scale to evaluate properly the symptoms occurring after drug disc
ontinuation is still under investigation. These studies will or should
permit the differentiation of rebound, withdrawal and recurrence. Met
hods developed to study reinforcing effects in post-addicts and health
y subjects are self-administration and choice procedures. In addition,
the more traditional approach has been through assessing self-reporte
d effects in which standardized questionnaires are used (Addiction Res
earch Center Inventory or A.R.C.I.; Single Dose Questionnaire or S.D.Q
.). A third focus of measurement has been discrimination studies perfo
rmed in individuals with histories of drug abuse as well as healthy su
bjects. Abuse-liability testing of a new compound needs a multidimensi
onal assessment to optimize the predictivity in defining the relative
risk.