PIP: Educational intervention programs in Florin, Greece, a mountainous rural area with a low (15-20%) attendance at prenatal clinics, were conducted to reduce perinatal and infant morbidity and mortality and to promote physical and psychomotor development. Prior community diagnostic surveys had identified low income, poor living conditions, and illiteracy as very closely linked with poor hygiene, poor nutrition, nonutilization of services, frequent infections and high perinatal and infant mortality. The objectives of the intervention were to assess the effects of health education on breast feeding practices and use of available medical services. 300 pregnant women participants were randomly identified by the clinic as the intervention group and 200 as controls. Both groups were similar with predominantly low socioeconomic status (69-73%) and peasant farmers. There were few basic housing amenities (13.1 with interventions and 12.7 for controls). 70% of the women lived in extended families. The intervention involved home visits on nutrition, general hygiene, breast feeding, and newborn care. Visits were scheduled every 2 weeks in the 1st 2 months of pregnancy and every month until the infant was 12 months old. It was found that nutrition counseling was positively associated with maternal weight between interventions and controls (11.33 vs. 10.30 p.05) but not on low birth weight. Prematurity was reduced (3.7% intervention vs. 8.3% controls, p.04). The perinatal mortality was 31% for interventions vs. 41% for controls, but there were a significantly higher number of fetal deaths (28 weeks) for interventions. There was no discernible impact on breast feeding practices except for demand feeding (61% interventions vs. 38% controls) perhaps because both groups considered breast milk the best. Overfeeding was affected by health counseling, but was reflected only after the 1st year (12.1% control vs. 6.7% intervention in the 90th percentile). The presence of anemia followed a similar pattern with intervention impact after the 1st year. There was no measurable impact in psychomotor development. Reported illness days had a significant lower mean 1 year for interventions. Illness reports checked against clinic visits showed underreporting for both groups. Colds, otitis, and gastroenteritis were the most common ailments. Both showed greater illness between 8 and 12 months with a slightly higher incidence of feverish episodes, gastroenteritis and otitis among controls. Hospitalizations were not significantly different. Both groups used the clinic for infant illness; home visits depressed the use of free routine checkups for interventions from 0 to 4 months. 5 neonatal deaths occurred among controls and interventions. It was anticipated that mortality is also affected by linkages to the community and improvement in medical services. The results were mixed but the program was partly successful.