Key points• The increased blood flow associated with local heating of the skin is ∼60% dependent on nitric oxide. The remaining ∼40% is unknown.• Endothelial-derived hyperpolarizing factors (EDHFs), a class of vasodilators, are known to contribute to increases in blood flow in other vascular beds.• In the present study, we showed the drug tetraethylammonium (which blocks the channels involved in EDHF-mediated vasodilatation), when given in combination with nitric oxide synthase inhibition, blocked the majority of hyperaemia to local heat, indicating that EDHFs are responsible for the majority of the remaining ∼40% of hyperaemia.• We also showed that about half of the EDHF-component is attributed to a specific type of EDHF, epoxyeicosatrienoic acid (EET), as evidenced using the cytochrome P450 inhibitor sulfaphenazole.• These findings help further our understanding of the mechanisms behind cutaneous thermal hyperaemia.Abstract While it is accepted that NO is responsible for ∼60% of the plateau in cutaneous thermal hyperaemia, a large portion of the response remains unknown. We sought to determine whether the remaining ∼40% could be attributed to EDHF-mediated activation of KCa channels, and whether the epoxyeicosatrienoic acids (EETs), derived via cytochrome P450, were the predominant EDHF active in the response. Four microdialysis fibres were placed in the forearm skin of 20 subjects. In Protocol 1 (n = 10): (1) Control, (2) N G -nitro-L-arginine methyl ester (L-NAME), (3) a KCa channel inhibitor, tetraethylammonium (TEA), and (4) TEA + L-NAME. In Protocol 2 (n = 10): (1) Control, (2) L-NAME, (3) a cytochrome P450 inhibitor, sulfaphenazole, and (4) sulfaphenazole + L-NAME. Local heating to 42• C was performed and skin blood flow was measured with laser Doppler flowmetry. Data are presented as the percentage of maximal cutaneous vascular conductance (CVC). All drug sites attenuated plateau CVC from the control site (86 ± 1%) to 79 ± 3% with sulfaphenazole (P = 0.02 from control), 71 ± 3% with TEA (P = 0.01 from control), and further to 38 ± 2% with L-NAME (P < 0.001 from control, P < 0.001 from TEA). Plateau was largely attenuated with sulfaphenazole + L-NAME (24 ± 2%; P = 0.002 from L-NAME), and nearly abolished with L-NAME + TEA (13 ± 2%; P = 0.001 from sulfaphenazole + L-NAME), which was not different from baseline (P = 0.14). Furthermore, the initial peak was just 17 ± 2% with TEA + L-NAME (P < 0.001 from L-NAME). These data suggest EDHFs are responsible for a large portion of initial peak and the remaining 40% of the