At the same post-menstrual age (39-41 weeks), EEG maturation assessed according to the Nolte and Haas method (Nolte, R. and Haas, H.G. (1978) Dev. Med. Child Neurol., 20, 167-182) was studied in 16 fullterm infants (FT), 17 healthy preterm infants (HP) and 18 pathological preterm infants (PP) affected by brain lesions (haemorrhage and/or leucomalacia). There were no significant differences in respect to EEG maturational codes, EEG types or bioelectrical age between the FT and HP groups. The preterm infants affected by brain lesions presented higher percentages of younger EEG codes (i.e. 36 weeks) in State 1 and a higher number of fluctuations between one maturation code and another in both State 1 and State 2, with respect to the HP group. Also, although the PP infants with young EEG codes did not present serious background EEG abnormalities, they did reveal minor EEG abnormalities, such as excessive asynchrony of the 'tracé alternant', lack of frontal sharp transients and monomorphic TA 'bouffees' with little activity at 2-6 c/s. However, no relationship between young EEG codes and onset-offset or duration of the states was found: young codes were often randomly distributed in successive State 1-State 2 epochs, regardless of groupings. Bioelectric age appropriate to the post-menstrual age precedes a normal development or only minor handicap at 24 months, while EEG immaturity of more than 2 weeks corresponds to later major handicaps. The prognostic value of EEG immaturity of between 1.1 and 2 weeks is uncertain.