Arrhythmic fluctuations in neural activity occur at many levels of the nervous system. Such activity does not have a characteristic temporal periodicity but can exhibit statistical similarities, most commonly power-law scaling behavior, which is indicative of scale-free dynamics. The recurrence of scaling laws across many different systems and its manifestation in behavior has prompted a search for unifying principles in human brain function. With this in mind, a focused search for abnormities in scale-free dynamics is of considerable clinical relevance to migraine and other clinical pain disorders. Here we examined the scale-free properties of the resting-state fMRI signal in the broadband frequency range known to be related to spontaneous neural activity (0.01-0.1Hz). In a large cohort of episodic migraine patients (N=40), we observed that the strength of long-range temporal correlations in the fMRI signal (captured by the scaling exponent α) was significantly higher in the sensorimotor network compared to healthy controls. Increases in the scaling exponent were positively correlated with fMRI signal variance and negatively correlated with patient's self-reported headache intensity. These changes in the fMRI signal suggest that the temporal structure of amplitude fluctuations carries valuable information about the dynamic state of the underlying neuronal networks and ensuing sensory impairments in migraine. The demonstrated scaling laws pose a novel quantitative approach for examining clinically relevant inter-individual variability in migraine and other pain disorders.