Background and purpose For the diagnosis of relapsing–remitting multiple sclerosis (RRMS), the revised 2017 McDonald criteria include cerebrospinal fluid specific oligoclonal bands as a new criterion for dissemination in time. Amongst other things, one expectation of the new criteria is to marginalize the diagnosis of clinically isolated syndrome (CIS), thus allowing for a faster and at the same time still reliable diagnosis of RRMS. Methods In this study, data from an unselected patient cohort with a typical CIS and dissemination in space at a large German Multiple Sclerosis Center from 2013 to 2016 were re‐analysed to compare differences in diagnosing RRMS with the 2017 versus 2010 McDonald criteria in everyday practice. Results Out of a cohort of 290 patients presenting with a typical first demyelinating event, 52% (152 patients) with the diagnosis of RRMS and 48% (138 patients) with the diagnosis of CIS according to the 2010 McDonald criteria were identified. The application of the 2017 McDonald criteria in the same patients increased the number of definite RRMS to 94% (273), thus leaving only 6% of patients with the diagnosis of CIS. The reason for this shift was the presence of cerebrospinal fluid specific oligoclonal bands which was found in 92.7% of the total population and in all patients with 2017 McDonald RRMS. Over a mean follow‐up of 1.5 years, 50% of patients formerly diagnosed with CIS who are now RRMS also fulfilled the 2010 McDonald criteria. Conclusions Our data support the use of the 2017 McDonald criteria for a more sensitive, but not that specific, diagnosis of RRMS in everyday practice.