The regulation of blood flow to any organ results from a complex regulation from several mechanisms, both central and local. Several mathematical tools have contributed to the better understanding the interaction between these mechanisms by exploring more detailed features of perfusion signals. Fractality is one of such features and can be defined as the self-similarity of a signal at different scales. Photoplethysmography (PPG) is a non-invasive and low-cost technology that allows the recording of skin/muscle perfusion over time. Previous publications have shown that PPG signals result from the contribution of the different physiological activities that affect perfusion (cardiac, respiration, myogenic, sympathetic, endothelial), each with its specific frequency interval. This study aimed at characterizing the fractal behaviour of PPG signals obtained during a suprasystolic limb occlusion (SLO) protocol. Ten healthy male subjects (mean 20.2 ± 2.3 y.o.) participated in this study after giving informed written consent. After acclimatization, subjects performed a standard SLO protocol on a random upper limb while sitting upright, as follows: 10 min resting with both arms at heart level (phase I), 5 min random arm occlusion (200 mmHg, phase II) and 10 min recovery in the initial position (phase III). PPG signals recorded from the index finger of both occluded and non-occluded limbs. These signals were first decomposed into their respective frequencies with the wavelet transform. Then, both the raw signal and the components were processed with a detrended fluctuation analysis (DFA) algorithm and the alpha exponent was calculated for each phase. The Wilcoxon signed rank test was used for comparing the alpha exponents between phases and the Mann-Whitney test for independent samples for limb comparisons (p<0.05). The magnitude of alpha exponents increased with decreasing frequency of the PPG components. Furthermore, occlusion significantly changed the alpha exponents of the PPG signal and several of its components from the occluded limb. These results show that PPG perfusion signals exhibit fractal behaviour and that the DFA-derived alpha exponent could serve as new descriptor of perfusion regulation phenomena.