The heterogeneity found in fibromyalgia (FM) patients has led to the investigation of disease subgroups, mainly based on clinical features. The aim of this study was to test the hypothesis that clinical FM subgroups are associated with different underlying pathophysiological mechanisms. Sixty-three FM patients were classified in type I or type II, according to the Fibromyalgia Impact Questionnaire (FIQ), and in mild/moderate versus severe FM, according to the severity of three cardinal symptoms considered in the American College of Rheumatology (ACR) 2010 criteria (unrefreshed sleep, cognitive problems and fatigue). To validate the subgroups obtained by these two classifications, we calculated the area under the receiver operating characteristic curves for various clinical variables and for two potential biomarkers of FM: Response to experimental pressure pain (algometry) and the amplitude/intensity slopes of the auditory evoked potentials (AEPs) obtained to stimuli of increasing intensity. The variables that best discriminated type I versus type II were those related to depression, while the indices of clinical or experimental pain (threshold or tolerance) did not significantly differ between them. The variables that best discriminated the mild/moderate versus severe subgroups were those related to the algometry. The AEPs did not allow discrimination among the generated subsets. The FIQ-based classification allows the identification of subgroups that differ in psychological distress, while the index based on the ACR 2010 criteria seems to be useful to characterize the severity of FM mainly based on hyperalgesia. The incorporation of potential biomarkers to generate or validate classification criteria is crucial to advance in the knowledge of FM and in the understanding of pathophysiological pathways.