The diagnostic criteria used in primary progressive (PP) and relapsing-remitting (RR) multiple sclerosis (MS) show substantial differences. This introduces complexity in the diagnosis of MS which could be resolved if these criteria could be unified in terms of the requirements for dissemination in space (DIS). The aim of this study was to assess whether a single algorithm may be used to demonstrate DIS in all forms of MS. Five sets of RRMS criteria for DIS were applied to a cohort of 145 patients with established PPMS (mean disease duration: 11 years - PPMS-1): C1: Barkhof-Tintoré (as in 2005 McDonald's criteria); C2: Swanton et al. (as in JNNP 2006); C3: presence of oligoclonal bands plus two lesions (as in McDonald's criteria); C4 and C5: a two-step approach was also followed (patients not fulfilling C1 or C2 were then assessed for C3). Two sets of PPMS criteria for DIS were applied: C6: Thompson et al. (as in 2001 McDonald's criteria); C7: 2005 McDonald criteria. A second sample of 55 patients with less than 5 years of disease duration (PPMS-2) was also analysed using an identical approach. For PPMS-1/PPMS-2, fulfilment was: C1:73.8%/66.7%; C2:72.1%/59.3%; C3:89%/79.2%; C4:96%/92.3%; C5:96%/85.7%; C6:85.8%/78.7%; C7:91%/80.4%. Levels of fulfilment suggest that the use of a single set of criteria for DIS in RRMS and PPMS might be feasible, and reinforce the added value of cerebrospinal fluid (CSF) findings to increase fulfilment in PPMS. Unification of the DIS criteria for both RRMS and PPMS could be considered in further revisions of the MS diagnostic criteria.