During the perioperative period, evaluation of digital blood flow would be useful in early detection of decreased circulating volume, thermoregulatory responses or anaphylactoid reactions, and assessment of the effects of vasoactive agents. This study was designed to assess the reliability of fingertip temperature, core-fingertip temperature gradients and fingertip-forearm temperature gradients as indices of fingertip blood flow in the clinical setting of the operating theatre. In 22 adult patients undergoing abdominal surgery with general anaesthesia, fingertip skin-surface temperature, forearm skin-surface temperature, and nasopharyngeal temperature were measured every five minutes during the surgery. Fingertip skin-surface blood flow was simultaneously estimated using laser Doppler flowmetry. These measurements were made in the same upper limb with an IV catheter (+ IV group, n=11) or without an IV catheter (-IV group, n=11). Fingertip blood flow, transformed to a logarithmic scale, significantly correlated with any of the three thermal measures in both the groups. Their rank order as an index of fingertip blood flow in the -IV group was forearm-fingertip temperature gradient (r=-0.86) > fingertip temperature (r=0.83) > nasopharyngeal-fingertip temperature gradient (r=-0.82), while that in the +IV group was nasopharyngeal-fingertip temperature gradient (r=-0.77) > fingertip temperature (r=0.71) > forearm-fingertip temperature gradient (r=-0.66). The relation of fingertip blood flow to each thermal measure in the -IV/group was stronger (P<0.05) than that in the +IV group. In the clinical setting of the operating theatre, using the upper limb without IV catheters, fingertip skin-surface temperature, nasopharyngeal-fingertip temperature gradients, and forearm-fingertip temperature gradients are almost equally reliable measures of fingertip skin-surface blood flow.