Background. Disability in multiple sclerosis (MS) is related to neuroaxonal degeneration. A reliable blood biomarker for neuroaxonal degeneration is needed. Objectives. To explore the relationship between cerebrospinal fluid (CSF) and serum concentrations of a protein biomarker for neuroaxonal degeneration, the neurofilaments heavy chain (NfH). Methods. An exploratory cross-sectional () and longitudinal () study on cerebrospinal fluid (CSF) and serum NfH phosphoform levels in patients with MS. The expanded disability status scale (EDSS), CSF, and serum levels of NfH-SMI34 and NfH-SMI35 were quantified at baseline. Disability progression was assessed at 3-year followup. Results. At baseline, patients with primary progressive MS (PPMS, EDSS 6) and secondary progressive MS (SPMS, EDSS 6) were more disabled compared to patients with relapsing remitting MS (RRMS, EDSS 2, ). Serum and CSF NfH phosphoform levels were not correlated. Baseline serum levels of the NfH-SMI34 were significantly () higher in patients with PPMS (2.05 ng/mL) compared to SPMS (0.03 ng/mL) and RRMS (1.56 ng/mL). In SPMS higher serum than CSF NfH-SMI34 levels predicted disability progression from baseline ( 2, ). In RRMS higher CSF than serum NfH-SMI35 levels predicted disability progression ( 2, ). Conclusion. Serum and CSF NfH-SMI34 and NfH-SMI35 levels did not correlate with each other in MS. The quantitative relationship of CSF and serum NfH levels suggests that neuroaxonal degeneration of the central nervous system is the likely cause for disability progression in RRMS. In more severely disabled patients with PP/SPMS, subtle pathology of the peripheral nervous system cannot be excluded as an alternative source for blood NfH levels. Therefore, the interpretation of blood protein biomarker data in diseases of the central nervous system (CNS) should consider the possibility that pathology of the peripheral nervous system (PNS) may influence the results.