Although drinking behavior is clearly a central dependent variable in
alcoholism treatment research, the field has reached no consensus on m
easurement methodology for alcohol consumption. At least four methods
for quantifying consumption have been commonly used in outcome studies
: quantity-frequency questions, average consumption grids, timeline fo
llow-back and self-monitoring. The Form 90 family of structured interv
iews was developed by collaboration among the Project MATCH investigat
ors, combining the strengths of prior assessment methodologies. The de
velopment, structure, supporting software and training approaches for
the Form 90 instruments are described.