Laser‐Doppler flowmetry (LDF) quantifies changes in localized microvascular perfusion in response to thermal and non‐thermal stimuli. Despite its widespread use, the reliability of the measurement during whole‐body heating is unknown. This presents a key knowledge gap that is critical for experimental design, statistical analysis, and interpretation. The purpose of this study was therefore to assess the reliability of LDF‐derived cutaneous vasodilation between separate days (between‐day) and between forearm sites (within‐day) during incremental whole‐body heat stress. On three occasions (~1 week apart), 11 healthy men (25 [SD 5] years) were passively heated (whole‐body heating using a water perfused suit) to increase esophageal temperature (Teso) by +0.7°C [0.11] (low heat stress, LHS) and +1.3°C [0.14] (moderate heat stress, MHS). Forearm skin blood flow (arbitrary perfusion units; PU) was measured in three adjacent sites on the mid‐dorsal forearm using LDF (Perimed) while local skin temperature was clamped (34°C). Maximal vasodilation was achieved by heating each skin site to 44°C (for ≥ 20 min), but without the use of SNP, while Teso was clamped at +1.8°C [0.1]. Beat‐to‐beat mean arterial pressure (MAP) was measured continuously. Cutaneous vascular conductance (CVC) at each skin site was calculated as PU/MAP. Data were presented as absolute CVC and percentage of maximal vasodilation (%CVC44). Reliability was assessed at baseline (no heat stress, NHS), LHS, and MHS using the intra‐class correlation coefficient (ICC), with values ≥0.7 considered acceptable for research purposes. Between days, CVC increased from 0.3 [0.2] to 2.0 PU [0.8] at LHS, and then to 2.1 PU [0.8] at MHS. Expressed as %CVC44 this equated to increases from 12 [5] to 68% [15] and then 73% [13]. Within‐day (i.e., between sites), average CVC increased from 0.3 [0.2] to 1.9 PU [0.7] at LHS, and 2.0 PU [0.7] at MHS, which related to an increase from 13 [6] to 71% [15] at LHS, and to 76% [12] at MHS when expressed as %CVC44. At NHS, the between‐ and within‐day reliability of these measurements did not exceed the 0.7 ICC threshold when expressed as CVC or %CVC44 at NHS (range: 0.00 to 0.45). Similarly, at LHS and MHS the between‐day reliability did not exceed 0.7 when expressed as CVC (0.57 and 0.69, respectively) or %CVC44 (both 0.57). While within‐day reliability of %CVC44 at LHS and MHS was not acceptable (0.54 and 0.64, respectively), it was when expressed as CVC (0.74 and 0.71, respectively). We show that the between‐day reliability of this technique was poor, irrespective of the level of whole‐body heat stress, but within‐day reliability was acceptable during heating, but not for resting measurements. This has important implications for studies assessing reactivity of skin vasculature to passive whole‐body hyperthermia, particularly with respect to the expected difference between the signal (i.e., expected change) and noise (i.e., measurement and biological variability), and the need to ensure that repeated measurements reflect the same microvascular environment.Support or Funding InformationNatural Sciences and Engineering Research Council of Canada