We investigated various magnetic resonance MRI parameters for both brain and spinal cord to see if any improved the clinicoradiological correlation in multiple sclerosis. Ninety-one multiple sclerosis patients (28 relapsing-remitting, 32 secondary progressive and 31 primary progressive) were imaged using conventional T1, proton density- and T2-weighted MRI of the brain and spinal cord. Focal brain and spinal cord lesion load was scored, as were diffuse signal abnormalities, brain ventricular volume and spinal cord cross-sectional area. Clinical measures included the expanded disability status scale (EDSS), the functional systems score and a dedicated urology complaint questionnaire. Secondary progressive patients differed from relapsing-remitting and primary progressive patients by a larger number of hypointense T1 lesions in the brain, ventricular enlargement and spinal cord atrophy. Primary progressive patients more often had diffuse abnormalities in the brain and/or spinal cord than did relapsing-remitting and secondary progressive patients. In the entire study population, EDSS correlated with both brain and spinal cord MRI parameters, which were independent. The urological complaint score correlated only with spinal cord MRI parameters. In relapsing-remitting and secondary progressive multiple sclerosis, the correlation between MRI and clinical parameters was better than in the entire population. In this subgroup EDSS variance could be explained best by T1 brain lesion load, ventricle volume and spinal cord cross-sectional area. In the primary progressive subgroup the clinicoradiological correlation was weak for brain parameters but was present between spinal cord symptoms and spinal cord MRI parameters. In conclusion, the different brain and spinal cord MRI parameters currently available revealed considerable heterogeneity between clinical subtypes of multiple sclerosis. In relapsing-remitting and secondary progressive multiple sclerosis both brain and spinal cord MRI may provide a tool for monitoring patients, while in primary progressive multiple sclerosis the clinicoradiological correlation is weak for brain imaging.