Objective To examine the association between gestational age, telomere length (TL) and rate of shortening in newborns. Study design Genomic DNA was isolated from buccal samples of 39 term infants at birth and one year and 32 preterm infants at birth, term-adjusted age (40 weeks post-conception) and age one-year corrected for gestational duration. Telomere length was measured by quantitative real-time PCR. Demographic and clinical data were collected during clinic or research visits and from hospital records. Socioeconomic status was estimated using the deprivation category (DEPCAT) scores derived from the Carstairs score of the subject’s postal code. Results At birth, preterm infants had longer telomeres than infants born at term. However, there was no difference in telomere length between preterm infants and term infants at one year of age, implying that the rate of telomere shortening was greater in pre-term than term infants. Interestingly, TL at age 40 weeks post-conception in preterm infants was significantly longer than term infant TL at birth, suggesting that time since conception is not the only factor that affects rate of shortening. Several factors, including sex, fetal growth restriction, maternal age, maternal booking body mass index (BMI), mother education level and DEPCAT score, also differed between the preterm and term groups. Conclusions Preterm infants have longer telomeres than term infants at birth. In the studied cohort, the rate of telomere shortening was greater in the premature group compared with the term infants. This finding agrees with previous studies using cord blood, suggesting that the longer TL in premature infants detected at birth do not persist and demonstrating that use of saliva DNA is acceptable for studies of telomere dynamics in infants. However, that the TL at age 40 weeks post-conception in preterm is longer than term infants at birth suggests that biological factors other than time since conception also affect rate of shortening.