Hypnosis may play an important role in reducing preterm labour for patients who have higher levels of psychosocial stress. This study examines the rate of late-preterm birth in a hypnosis group (directed to all women) and a historical control group. From July 2007 all women (n = 64), who were in their 28th to 34th weeks' gestation, were offered self-hypnosis training using the hypnoreflexogenous protocol after Hüsken-Janßen and Schauble. Expectant mothers with uncertain anticipated days of delivery were excluded. All women who delivered after 31 weeks' gestation served as a control group (n = 2135) from January 2006 till June 2007. In the hypnosis group there were three preterm deliveries (4.7%) (before 37 + 0 weeks' gestation) whereas in the control group there were 220 preterm deliveries (10.3%) (p = 0.01). Average cigarette usage during the current pregnancy was lower in the hypnosis group (p = 0.02). Higher work-educated employments (p = 0.01), higher age of the mother (p < 0.001) and fewer previous pregnancies (p < 0.03) were found in the hypnosis group. Preterm birth correlated with the number of previous pregnancies (−0.38; p < 0.001) but not with smoking. Hypnosis was shown to be effective therapy without side-effects, which can reduce preterm delivery. This clinical study showed a significant prevention of preterm delivery. Prospective randomized controlled studies are required to evaluate fully the preventive value of clinical hypnosis. Copyright © 2009 British Society of Experimental & Clinical Hypnosis. Published by John Wiley & Sons, Ltd.