Objectives: Many people affected by multiple sclerosis (MS) experience cognitive impairment, especially decreases in information processing speed (PS). Neural disconnection is thought to represent the neural marker of this symptom, although the role played by alterations of specific functional brain networks still remains unclear. The aim is to investigate and compare patterns of association between PS-demanding cognitive performance and functional connectivity across two MS phenotypes. Methods: Forty patients with relapsing-remitting MS (RRMS) and twenty-five with secondary progressive MS (SPMS) had neuropsychological and MRI assessments. Multiple regression models were used to investigate the relationship between performance on tests of visuo-motor and verbal PS, and on the verbal fluency tests, and functional connectivity of four cognitive networks, i.e. left and right fronto-parietal, salience and default-mode, and two control networks, i.e. visual and sensorimotor. Results: Patients with SPMS were older and had longer disease history than patients with RRMS and presented with worse overall clinical conditions: higher disease severity, total lesion volume and cognitive impairment rates. However, in both patient samples, cognitive performance across tests was negatively correlated with functional connectivity of the salience and default-mode networks, and positively with connectivity of the left frontoparietal network. Only the visuo-motor PS scores of the RRMS group were also associated with connectivity of the sensorimotor network. Conclusions: PS-demanding cognitive performance in patients with MS appears mainly associated with strength of functional connectivity of frontal networks involved in evaluation and manipulation of information, as well as the default mode network. These results appear in line with the hypothesis that multiple neural networks may be needed to support normal 4 cognitive performance across MS phenotypes. However, different PS measures showed partially different patterns of association with functional connectivity. Therefore, further investigations are needed to clarify the contribution of inter-network communication to specific cognitive deficits due to MS.