The aim of the study was to assess if lung function at birth predicts lung function at 2 yr and secondly, if lung function development was influenced by the common phenotypes of recurrent bronchial obstruction (rBO) or atopic eczema (AE) by 2 yr. Lung function was assessed at birth (n = 802) and at 2 yr within the prospective birth cohort study 'the Environment and Childhood Asthma Study' in Oslo. The 135 children with lung function measured at birth by tidal flow volume (TFV) loops and passive respiratory mechanics, who were included in a nested case-control study were reinvestigated at 2 yr with clinical examination, TFV loops (n = 90) (mean age 26.6 (3.7 s.d.) months), skin prick test and parental interview. Children were categorized into quartiles (lower, middle two, upper) according to time to peak tidal expiratory flow/total expiratory time (t(PTEF)/t(E)) at birth as well as clinical phenotype based on the presence of rBO and/or AE (ever) by 2 yr. The observed reduction in mean t(PTEF)/t(E) from birth to 2 yr within the quartiles, were not significantly different after controlling for 'regression to the mean'. t(PTEF)/t(E) at birth correlated significantly with t(PTEF)/t(E) at 2 yr, (r = 0.475, p < 0.001). Children with both rBO and AE by 2 yr had significantly lower t(PTEF)/t(E) at 2 yr (p = 0.002) and at birth (p = 0.027), compared with children with no rBO or AE. Clinical phenotype at 2 yr did not influence the change in t(PTEF)/t(E) from birth to 2 yr. This study demonstrates a clear tracking of lung function from birth, not influenced by rBO or AE by 2 yr.