The benefits of early treatment of nutritional and respiratory problems in the CF infant and of genetic counselling for the parents are widely recognized. However, clinical diagnosis of CF is often delayed despite early onset of symptoms and the usefulness of neonatal population screening as a preventive measure is still under debate. This study analyses the clinical history of CF patients diagnosed exclusively on the basis of positive neonatal screening tests with the aim of identifying the earliest markers of the disease. We studied 103 CF infants born in north-east Italy, diagnosed following neonatal screening: assay of immunoreactive trypsin (IRT) from a heel-prick blood sample followed by a measurement of meconium lactase in cases with raised IRT. Diagnosis was confirmed by sweat test at an average age of 39 days. Eighty-one patients (79%) had symptoms strongly suggestive of CF at diagnosis, and signs and/or symptoms of pancreatic insufficiency were present in 16 of the remaining 22 cases. The most frequent symptom was growth failure (69% of infants) and of these, 44% weighed the same as at birth or less. Pancreatic insufficiency was confirmed by the low level of faecal chymotrypsin found in 85% of cases. IRT was elevated in all cases. CF had not been suspected in any symptomatic infant, although most of the infants had been monitored by a paediatrician. In conclusion, most infants with CF diagnosed by neonatal screening are already symptomatic in the first six weeks of life and the most frequent symptom is failure to thrive; pancreatic insufficiency was already present in most cases. In areas without CF neonatal screening programs, the disease should be excluded by differential diagnosis in all cases with growth failure notwithstanding adequate caloric intake in the first months of life. The high sensitivity, low cost and simple execution of IRT and fecal chymotrypsin tests make them an ideal first step in suspect cases before proceeding to the sweat test, often performed late because of limited availability.