The Primary Mental Health Projcct (I'MHP), a program for early identification and prevention of school maladaptation, began 17 ycars ago. The project's initial goals were to devclop accurate, mass-screening procedures for identifying school adjustment problems, and programs to help children to adapt more effectively (Cowen, Izzo, Miles, Telschow, Trost and Zax, 1963;Cowen, Zax, Izzo, and Trost, 1966). I n 1'MHl''s second phase, housewives were recruited and trained as child-aides to work with early-identified maladapting children (Zax, Cowen, Izzo, Rladonia, Merenda, and Trost, 1966;Zax and Cowen, 1967). In its third, most recent stage, less well-resourced extensions of P M H P have been extended to a number of schools in four urban and suburban districts (Cowen, Dorr, Izzo, Madonia, and Trost, 1971). P M H P has emphasized research from its start, and has reported a number of outcornc studies-broad and specific-in its 17 year history. Evaluations of the original PMHP werc broad and included all project children-thc adjusted as well as the vulnerable. These studies also used a variety of assessment criteria: e.q., school rccord measures, teacher reports, self-reports, achievemcnt indexes and peer evaluations (Cowen, et aZ., 1963(Cowen, et aZ., , 1966. Outcome studies done during PMHP's second (aidc) stage were restricted to aide-seen childrcn and had fewer criterion measures. There havc been two part-evaluations of PMH1"s third (expansion) stagc, which began in 1969: an aide evaluation report (Dorr and Cowen, 1973) and a comparison of project ''terminators'' us. "non-terminators" (Cowen and Schochet, 1973).The findings of PMHI' outcome studies are reasonably consistent. The initial series (e.g., Cowen, et al., 19GG) showed that, after three years, project children werc doing hettcr than controls on a range of school rccord, pcrformance, and adjustment measures. In phase two, childrcn seen by aides showed both short term (Cowen, 1968) and long term gains (Cowen, Dorr, Trost, and Izzo, 1972). Studies based on the currcnt expanded program indicatc that the aides judged positive outcomes to havc takcn place (Dorr and Cowcn, 1973), and that successful I'MH1' "terminators" improved significari tly morc than "non terminators" on teacher and aide behavior-rating measures (Cowen and Schochct, 1973).Rarlicr PMHI' outcome cvaluations, comparcd to recent ones, werc bctter controllcd, morc rigorous, and lcss encuinbcrcd by cxtraneous vnriablcs. They also used broadcr evaluation criteria and wcre based on broader samples of children, including well-functioning oiics. Thus, the early project's circumscribed nature allowed for tighter evaluation. Itcscarch rigor, in cvaluatirig PMHI', has recently weakened, due primarily to thc complcx realities of today's rapidly changing educational scenc and the rnaiiy problems involvcd in moving from a narrow, singleschool rescarch base to :I complex program in 13 schools in four individualized 'This st,iidy w.zs cnrried orit, under grant support froin the NIMII ICxperimental and Speci:L...