We tested the hypothesis that microdialysis of hypertonic saline would attenuate the skin blood flow response to local heating. Seventeen healthy subjects (23 ± 1 years old) were studied. In one group (n = 9), four microdialysis fibres were placed in the forearm skin and infused with the following: (1) Ringer solution; (2) normal saline (0.9% NaCl); (3) hypertonic saline (3% NaCl); and (4) 10 mm l-NAME. A second group (n = 8) was infused with the following: (1) normal saline; (2) hypertonic saline; (3) normal saline + l-NAME; and (4) hypertonic saline + l-NAME. Red blood cell flux was measured via laser Doppler flowmetry during local heating to 42°C. Site-specific maximal vasodilatation was determined by infusing 28 mm sodium nitroprusside while the skin was heated to 43°C. Data were expressed as the percentage of maximal cutaneous vascular conductance (%CVC(max)). The local heating response at the Ringer solution and normal saline sites did not differ (n = 9; initial peak Ringer solution, 69 ± 6 versus normal saline, 66 ± 2%CVC(max); plateau Ringer solution, 89 ± 4 versus normal saline, 89 ± 5%CVC(max)). Hypertonic saline reduced the initial peak (n = 9; normal saline, 66 ± 2 versus hypertonic saline, 54 ± 4%CVC(max); P < 0.05) and plateau (normal saline, 89 ± 5 versus hypertonic saline, 78 ± 2%CVC(max); P < 0.05) compared with normal saline. Plateau %CVC(max) was attenuated to a similar value at the normal saline + l-NAME and hypertonic saline + l-NAME sites (n = 8; normal saline + l-NAME, 39 ± 6 and hypertonic saline + l-NAME, 39 ± 5%CVC(max)). The nitric oxide contribution (plateau %CVC(max) - l-NAME plateau %CVC(max)) was lower at the hypertonic saline site (normal saline, 55 ± 6 versus hypertonic saline, 35 ± 4; P < 0.01). These data suggest an effect of salt on the cutaneous response to local heating, which may be mediated through a decreased production and/or availability of nitric oxide.