The high incidence of preterm birth (<37 weeks gestation) is a major concern in assisted reproductive technology. The objective of this study was to compare the risk of preterm birth from singleton pregnancies following either low technology treatment (intrauterine insemination and donor insemination) or high technology treatment (IVF, ICSI and gamete intra-Fallopian transfer) with that of naturally conceived pregnancies. Three cohorts of pregnancies resulting from either low or high technology treatment or from natural conception were included in the study. A number of potential risk factors were adjusted for. The incidence of very preterm birth (<32 weeks of gestation) was not significantly increased in the low technology treatment group (1.0 versus 1.3% in controls) but was significantly higher in the high technology treatment group (5.2%, P < 0.001). In spontaneous, elective Caesarean section (CS) and induced delivery onset, the risk of preterm birth increased gradually from the controls to the low technology treatment group to the high technology treatment group, while for an emergency CS the risk of preterm birth was very high in both treatment groups. The overall incidence of preterm birth increased significantly from the controls to the low technology treatment group and to the high technology treatment group. Logistic regression analysis showed that younger and older age, previous perinatal death and emergency CS were associated with an increased risk, while a previous live birth reduced the risk. The length of the infertile period did not seem to affect the risk in any of the treatment groups.