Data were collected longitudinally on 48 families qf returned prisoners of war to identify the best combination of factors to be used to explain the degree of reintegration of the returnee into his family system. The first year qf fbllow-up data was reported and four sets of data were considered: (1) background characteristics of husband and wife; (2) indices qf family preparedness for separation and reunion; (3) reports of returnees' prison experiences and their psychiatric status at repatriation; (4) measures qf family adjustment during separation. Linear multiple regression procedures were utilized to analyze the contribution of each predictor variable in accounting fbr the variance in the criterion and fbr deriving an equation where variables could be optimally weighted.