We analyzed the relationship between normal neonatal EEG features and gestational age and conceptional age, and evaluated the normal aspects of EEG maturation in preterm babies compared to term babies. We report 46 newborns, divided into two groups. Group I consisted of 11 newborns with gestational age between 30 and 32 weeks, followed with weekly polysomnograms until they reach 42 weeks' conceptional age. Group II (control) consisted of 35 newborns with gestational ages of 34 weeks (n = 5), 36 weeks (n = 10), 38 weeks (n = 10) and 40 weeks (n = 10) evaluated with one polysomnogram in their first 24-48 h of life. In each examination one 5 min epoch in REM and NREM sleep was analyzed to quantify the number of delta brushes, the presence of frontal and temporal sharp transients, the presence of delta frontal rhythmic activity, the grade of concordance between EEG patterns and sleep stages, the percent of interhemispheric synchrony and the duration of interburst interval. The age dependent variability of the EEG patterns was evaluated during the subsequent weeks with group comparisons at weeks 34, 36, 38 and 40. Our results show that the number of delta brushes and the duration of the interburst interval decrease as gestational and conceptional age increase. The percent of interhemispheric synchrony increases with gestational and conceptional age. The presence of frontal sharp transients and delta frontal rhythmic activity suggest that the newborn is fullterm. The presence of temporal sharp transients suggest a preterm newborn. The degree of concordance between behavioral sleep patterns and EEG was more helpful in recognizing sleep stages than in estimating gestational or conceptional age. Although the EEG patterns were comparable between the groups at the same age, analyses of the behavioral patterns of concordance in NREM sleep showed that newborns in Group I had a more immature behaviour than newborns in Group II. Our results also suggest that extrauterine life of preterm babies does not seem to accelerate EEG maturation but may influence the acquisition of behavioral patterns during NREM sleep.