The present study extends an on-going approach to preventative intervention based on translating findings from basic research on relations between marital discord and adolescent adjustment into brief (4-visit) psychoeducational programs. One such intervention program, The Family Communication Project (FCP), is evaluated in terms of its aims of improving family-wide, parent-adolescent, and adolescent adjustment. Building off of previous studies of the program's efficacy, we evaluated the program's longitudinal (3-year) effects on observed family-wide conflict behaviors, parenting behaviors, family reports of the family environment, parent-adolescent communication, and adolescent psychological adjustment. We further investigated potential theory-based mechanisms of change – including Emotional Security Theory (EST; Davies & Cummings, 1994) - by which improvements in family-wide conflict may lead to consequent improvements in family and adolescent adjustment. Participants included 225 couples with a participating adolescent child between 11 and 17 years of age. Families were randomly assigned to groups: (1) a parent-only group (n = 75), (2) a parent-adolescent group (n = 70), or (3) a non-intervention group (n = 80). Pretest, post-test, 1-year, and 3-year assessments were conducted. Structural Equation Modeling was used to test main effects of the intervention as well as research-informed process models of family conflict and its effects on later family and adolescent psychological adjustment. In addressing the three aims that we set out to test (i.e., the direct effects of the intervention, whether or not these effects occurred through changes in family conflict behaviors, and whether emotional insecurity mediates the effect of the intervention on child psychological outcomes), we found support for the efficacy of the intervention itself at improving several long-term outcomes (both observed and reported). Direct effects of group assignment on observed constructive conflict behaviors, maternal parenting attitudes, and family reports of the family environment were found; however, contrary to our hypothesis, these changes did not occur through observed changes in conflict behaviors. Instead, emotional insecurity mediated the effect of the intervention on adolescent psychological adjustment (externalizing symptoms) at the 3-year follow-up. Some support was found for the moderating effect of adolescent gender; no moderating effects of adolescent age were indicated.