The rate of preterm birth in the developed world has been shown to be increasing, in part attributable to obstetric intervention. It has been suggested that this may be a differential increase between socio-economic groups. We aimed to assess whether the preterm rate in Norway is different in socio-economic groups defined by maternal education, and to determine the extent to which a difference is attributable to a socio-economic differential in obstetrical intervention, in terms of caesarean section or induction of labour. We used data from the Medical Birth Registry of Norway from 1980 to 1998 with preterm rate as the outcome and maternal educational level, marital status and obstetric intervention as exposure variables. In multivariable analyses, adjustment was made for maternal age, year of birth and birth order, and secular trends were assessed according to year of birth. The preterm birth rate was highest in the lowest socio-economic group. An increase of 25.2% in the preterm rate was seen over the observation period. No apparent differential was seen in the increase of the crude preterm rates between socio-economic groups, although in multivariable analyses there was a significant interaction between socio-economic group and time, implying a stronger effect of low education towards the end of the observation period attributable to demographic change. In conclusion, the preterm birth rate increased over time, but was mainly due to an increase in obstetric interventions. No closing of the gap between socio-economic groups was observed.