This study investigates long-term effects on adult attachment due to temporary childhood separation as a result of the British evacuation of children during World War 2. A total of 859 respondents, aged 62-72 years, were recruited who had childhood homes in the county of Kent during the war. Of these, 770 had been evacuated and 89 remained at home and formed a non-evacuated control group. They participated in this retrospective survey of possible associations between childhood experiences of the evacuation, early upbringing, and later life-course variables, with adult attachment style assessed by the Relationship Questionnaire (Bartholomew & Horowitz, 1991). Reflecting the wartime concerns of Bowlby, male and female respondents evacuated between the ages of 4 and 6 years showed low incidences in secure attachment style of 38% and 27%, compared to those not evacuated at 64% and 44%, respectively, with a corresponding increase in the fearful category. The quality of care received during evacuation and the frequency of parental visits were also found to be significantly associated with attachment style for female respondents. Irrespective of experience of evacuation, both male and female respondents who reported poor home nurture showed a similar low incidence of secure attachment of 23% compared to those from homes with good quality care of 45% and 43%, respectively, with concomitant increases in proportions in the dismissing category for males and the fearful category for females. The clinical implications of the study are briefly discussed.