As a part of a prospective study on the psychological adjustment of Japanese teenagers enrolled in a foreign exchange student program, the Parental Bonding Instrument (PBI), the Maudsley Personality Inventory and the General Health Questionnaire (GHQ) were administered to 130 females and 47 males aged 17 through 19 before departure. Parental practices were found to influence personality features, which in turn contribute to the mental health status, but no direct significant relationship was found between the PBI and GHQ measures. The last finding is in contradistinction to the previous reports which report low care and high overprotection in unipolar depressives and anxiety disorder patients but in accordance with a recent finding which failed to replicate them in general population. It is possible that the earlier findings in clinical populations were a case of Berkson bias.