This paper underscores the important role Flayed by theory-based mediating
variables in randomized clinical trials. Indeed, it is essential that we kn
ow what mediating variables are relevant for particular outcomes in randomi
zed clinical trials and that we design interventions to optimize change in
the mediators of interest. Yet, knowledge of the pivotal mediating variable
s in behavior change does not imply that we know how to inter:ene effective
ly. This gap may be due to the fact that existing research typically has be
en designed to garner support for theory, as opposed to testing support for
behavior change strategies that are based on theory. In addition, we argue
that there are important mediating variables in behavior change that have
not been given systematic attention in theory development. For example, beh
avior change must be viewed as a collaborative process, and participants' p
erceptions concerning the feasibility of change is important to motivationa
l processes in randomized clinical trials. Control Clin Trials 2000;21:164S
-170S (C) Elsevier Science Inc. 2000.