Key points Over 15 million babies are born prematurely each year with approximately 1 million of these babies dying as a direct result of preterm delivery. β2‐Adrenoreceptor agonists that act via cAMP can reduce uterine contractions to delay preterm labour, but their ability to repress uterine contractions lasts ≤ 48 h and their use does not improve neonatal outcomes. Previous research has suggested that cAMP inhibits myometrial contractions via protein kinase A (PKA) activation, but this has yet to be demonstrated with PKA‐specific agonists. We investigated the role of PKA in mediating cAMP‐induced human myometrial relaxation, and the impact of prolonged cAMP elevation on myometrial contractility. Our findings suggest that PKA is not the sole mediator of cAMP‐induced myometrial relaxation and that prolonged prophylactic elevation of cAMP alone is unlikely to prevent preterm labour (PTL). Abstract Acute cAMP elevation inhibits myometrial contractility, but the mechanisms responsible are not fully elucidated and the long‐term effects are uncertain. Both need to be defined in pregnant human myometrium before the therapeutic potential of cAMP‐elevating agents in the prevention of preterm labour can be realised. In the present study, we tested the hypotheses that PKA activity is necessary for cAMP‐induced myometrial relaxation, and that prolonged cAMP elevation can prevent myometrial contractions. Myometrial tissues obtained from term, pre‐labour elective Caesarean sections were exposed to receptor‐independent cAMP agonists to determine the relationship between myometrial contractility (spontaneous and oxytocin‐induced), PKA activity, HSP20 phosphorylation and expression of contraction‐associated and cAMP signalling proteins. Acute (1 h) application of cAMP agonists promoted myometrial relaxation, but this was weakly related to PKA activation. A PKA‐specific activator, 6‐Bnz‐cAMP, increased PKA activity (6.8 ± 2.0 mean fold versus vehicle; P = 0.0313) without inducing myometrial relaxation. Spontaneous myometrial contractility declined after 24 h but was less marked when tissues were constantly exposed to cAMP agonists, especially for 8‐bromo‐cAMP (4.3 ± 1.2 mean fold versus vehicle; P = 0.0043); this was associated with changes to calponin, cofilin and HSP20 phosphorylated/total protein levels. Oxytocin‐induced contractions were unaffected by pre‐incubation with cAMP agonists despite treatments being able to enhance PKA activity and HSP20 phosphorylation. These data suggest that cAMP‐induced myometrial relaxation is not solely dependent on PKA activity and the ability of cAMP agonists to repress myometrial contractility is lost with prolonged exposure. We conclude that cAMP agonist treatment alone may not prevent preterm labour.