Objective: To summarize the advantages and limitations of general population, high-risk and "natural experiment" longitudinal studies for studying psychological change. The English and Romanian Adoptees study is used as an example of a "natural experiment," and detailed findings are provided. Method: What is new is a focus on the young people who spent the whole of their life in institutional care up to the time of adoption and who did not show subnutrition. The results were compared with a composite comparison group who had not experienced institutional care or who were adopted before the age of 6 months. The outcomes were assessed in terms of previously established deprivation-specific patterns (DSPs). Results: "Pure" psychosocial deprivation was associated with a substantial increase in the rate of DSPs. It was not associated with significantly impaired head growth if institutional care lasted less than 6 months, whereas thereafter there was a 2.5 standard deviation reduction. Subnutrition differed in being accompanied by impaired head and body growth even with institutional care lasting less than 6 months. In the pure psychosocial deprivation group, 45.5% showed a DSP at 15 years compared with 1.3% in the comparison group. Conclusion: "Pure" psychosocial deprivation (in the absence of subnutrition) had a profound effect on psychological functioning in the form of DSPs. Subnutrition had a surprisingly small effect on DSPs. (Contains 4 figures.)