Key points The role of prostaglandins in the changes in blood flow and microvascular vasodilation associated with exercise and muscle contraction is controversial. Whether prostaglandins are produced independently during muscle contraction or whether their production is dependent on the production of adenosine is not well understood. We show that prostaglandins are an important component of the microvascular vasodilation associated with muscle contraction but only under specific contractile conditions. Further, we show that microvascular vasodilation in response to adenosine is not dependent on prostaglandins. Therefore, we conclude that there are specific contractile conditions under which prostaglandins are an important component of the vasodilation induced by muscle contraction and we propose that prostaglandin‐induced microvascular vasodilation during exercise is independent of adenosine. Abstract Blood flow data from contracting muscle in humans indicates that adenosine (ADO) stimulates the production of nitric oxide (NO) and vasodilating prostaglandins (PG) to produce arteriolar vasodilatation in a redundant fashion such that when one is inhibited the other can compensate. We sought to determine whether these redundant mechanisms are employed at the microvascular level. First, we determined whether PGs were involved in active hyperaemia at the microvascular level. We stimulated four to five skeletal muscle fibres in the anaesthetized hamster cremaster preparation in situ and measured the change in diameter of 2A arterioles (maximum diameter 40 μm, third arteriolar level up from the capillaries) at a site of overlap with the stimulated muscle fibres before and after 2 min of contraction [stimulus frequencies: 4, 20 and 60 Hz at 15 contractions per minute (CPM) or contraction frequencies of 6, 15 or 60 CPM at 20 Hz; 250 ms train duration]. Muscle fibres were stimulated in the absence and presence of the phospholipase A2 inhibitor quinacrine. Further, we applied a range of concentrations of ADO (10−7–10−5 m) extraluminally, (to mimic muscle contraction) in the absence and presence of l‐NAME (NO synthase inhibitor), indomethacin (INDO, cyclooxygenase inhibitor) and l‐NAME + INDO and observed the response of 2A arterioles. We repeated the latter experiment on a different level of the cremaster microvasculature (1A arterioles) and on the microvasculature of a different skeletal muscle (gluteus maximus, 2A arterioles). We observed that quinacrine inhibited vasodilatation during muscle contraction at intermediate and high contraction frequencies (15 and 60 CPM). l‐NAME, INDO and l‐NAME + INDO were not effective at inhibiting vasodilatation induced by any concentration of ADO tested in 2A and 1A arterioles in the cremaster muscle or 2A arterioles in the gluteus maximus muscle. Our data show that PGs are involved in the vasodilatation of the microvasculature in response to muscle contraction but did not obtain evidence that extraluminal ADO causes vasodilatation through NO or PG or both. Thus, we propose that PG‐induced microvascular vasodilation during exercise is independent of ADO.