© 2018 Royal College of Obstetricians and Gynaecologists Objective: Determine associations of cardiorespiratory fitness, exercise systolic blood pressure (SBP) and heart rate recovery (HRR) following a maximal exercise test performed years preceding pregnancy with odds of preterm birth (PTB; <37 weeks' gestation) and small for gestational age (SGA; birthweight <10 th percentile) delivery. Design: Prospective, longitudinal. Setting: Multi-site, observational cohort study initially consisting of 2787 black and white women aged 18–30 at baseline (1985–86) and followed for 25 years (Y25; 2010–2011). Population: 768 nulliparous women at baseline who reported ≥1 live birth by the Y25 exam. Methods: We used Poisson regression to determine associations of exposures with PTB/SGA. Main outcome measures: PTB and/or SGA births. Results: Women with PTB (n = 143) and/or SGA (n = 88) were younger, had completed fewer years of education and were more likely to be black versus women without PTB/SGA (n = 546). Women with PTB/SGA had lower fitness (501 ± 9 versus 535 ± 6 seconds, P < 0.002) and higher submaximal SBP than women without PTB/SGA (144 ± 1 versus 142 ± 1 mmHg, P < 0.04). After adjustment, no exercise test variables were associated with PTB/SGA, though the association with HRR and submaximal SBP approached significance in the subset of women who completed the exercise test <5 years before the index birth. Conclusions: Neither fitness nor haemodynamic responses to exercise a median of 5 years preceding pregnancy, were associated with PTB/SGA. These findings indicate excess likelihood of PTB/SGA is not detectable by low fitness or exercise haemodynamic responses 5 years preceding pregnancy, but exercise testing, especially HRR and submaximal SBP, may be more useful when conducted closer to the onset of pregnancy. Tweetable abstract: Exercise testing conducted >5 years before pregnancy may not detect women likely to have PTB/SGA.