To differentiate between educationally subnormal students who do and who do not respond well to special school education, 161 children attending special classes in South Wales were studies. The first (the E-) group consisted of 47 children (mean IQ 55) found unsuitable for special school education. The second (the S-) group included 57 children (mean IQ 73) who had been transferred back to the ordinary school system and did well. Data obtained from interviews with mothers were analyzed; 35 hypotheses concerning congenital factors, early development, home standards, economic and material factors, family well-being, and cultural factors were tested. Results indicated that the E-group had a high incidence of organic damage, delayed development, and illness; the S-group tended to show normal early development but abnormal home environments. Conclusions drawn were that early identification of retardation is needed and that environmentally and organically handicapped children may require different educational programs. Six case studies are presented. (LE)