Pulse transit time (PTT) is a simple, noninvasive technique which shows great promise in the continuous monitoring of blood pressure and the assessment of arterial stiffness, and has potential applications in wearable healthcare devices. Usually, PTT is measured from the electrocardiogram (ECG) R-wave to a characteristic point on the peripheral pulse by photoplethysmography. However, peripheral blood circulation is sensitive to local temperature variation. This study investigated the effect of mild cold exposure on PTT by immersing one finger into cold water. The photoplethysmographic (PPG) signals were recorded from the cooled and adjacent uncooled fingers. The significant change in dc and ac amplitudes of the PPG pulse indicates that the mild cold exposure has a substantial effect on finger blood circulation. Three measurements of PTT were performed by calculating the time delay from the ECG R-wave to three different characteristic points on the PPG pulse, namely, the foot (PTT1), the point on the rising limb with maximal slope (PTT2) and the peak (PTT3), respectively. PTT3 was found to be easily subject to waveform distortion. Significant changes in the difference of PTT1 and PTT2 between two fingers were observed after a 5 min recovery period. The changes in PTT1 and PTT2 on the reference finger also showed close correlation (r = −0.77 and r = −0.80, p < 0.001) with the changes in SBP after recovery, whereas those measured on the test finger have low correlation (r = −0.53 and r = −0.38, p > 0.15). The results suggest that mild cold exposure may have a delay effect on PTT due to cold-induced vasodilatation and could be a potential source of error. The effect of cold exposure should be carefully examined, especially when PTT that includes a large portion of rising time of peripheral pulse is applied.