Longitudinal follow-up of preterm neonates after discharge from Intensive Care satisfies the family's need for support and ensures an ongoing link with the hospital concerned, but must be based on a specific, sensitive and predictive screening program. We report a prospective cohort study in a tertiary neonatal intensive care unit (NICU) using the Perinatal Risk Inventory (PERI)1, at discharge. Of 87 consecutive newborn referred to the tertiary NICU at the Pediatrics Department of Padua University in 1993 (gestational age > or = 25 weeks, weight at birth < 1500 g), 65 (75%) survived: 54 cases complied with the Fitzhardinge '80 criteria and were followed up to at least 1 year (corrected age). A further 18 neonates with neonatal cerebral problems were also considered. The PERI at discharge was combined with neonatological and neuro-evolutional evaluation and examination of neuro-sensorial development indicators at the corrected ages of 0, 3, 6 and 12 months. PERI scores were: < or = 7 in 36; 8-11 in 22; > 11 in 14. The cut off point at 7 to 11 increases the sensitivity, the specificity and the predictive values of the test. At 1 year (corrected age), 6 patients (8.3%) had a diagnosis of spastic diplegia and 2 (2.7%) had spastic hemiplegia. The relative risk of cerebral paralysis was 4.5 in the neonates with IVH 4 degrees, odds ratio 6.7, and 3.6 in the 16 patients with bronchopulmonary dysplasia (BPD), odds ratio 4.7. 72 patients were selected for follow-up by the Fitzhardinge '80 criteria, whereas with the PERI (cut off point 7) only 44 neonates would have been selected. This confirms the need for new criteria in recruiting preterm neonates for longitudinal follow-up to quantify neurological risks, particularly in view of the socioeconomic impact of the problem.