IntroductionPersonalized medicine is the holy grail of medicine. The EULAR recommendations for the management of rheumatoid arthritis (RA) support differential treatment between patients with baseline characteristics suggestive of a non-poor prognosis (non-PP) or poor prognosis (PP) (presence of autoantibodies, a high inflammatory activity and damage on radiographs). We aimed to determine which prognostic risk groups benefit more from initial monotherapy or initial combination therapy.Methods508 patients were randomized to initial monotherapy (iMono) or initial combination therapy (iCombo). Disease outcomes of iMono and iCombo were compared within non-PP or PP groups as determined on baseline characteristicsResultsPP patients treated with iCombo after three months more often achieved ACR20 (70% vs 38%, P <0.001), ACR50 (48% vs 13%, P <0.001) and ACR70 response (24% vs 4%, P <0.001) than those treated with iMono, and had more improvement in HAQ (median decrease 0.75 vs 0.38, P <0.001). After 1 year, differences in ACR20 response and DAS-remission remained; PP patients treated with iCombo (vs iMono) had less radiographic progression (median 0.0 vs 1.5, P =0.001).Non-PP patients treated with iCombo after three months more often achieved an ACR response (ACR20: 71% versus 44%, P <0.001; ACR50: 49% vs 13%, P <0.001; ACR70: 17% vs 3%, P =0.001) than with iMono, and functional ability showed greater improvement (median decrease in HAQ 0.63 vs 0.38, P <0.001). After 1 year, differences in ACR20 and ACR50 response remained; radiographic progression was comparable between the groups.Non-PP and PP patients responded equally well to iCombo in terms of improvement of functional ability, with similar toxicity.ConclusionsSince PP and non-PP patients benefit equally from iCombo through earlier clinical response and functional improvement than with iMono, we conclude that personalized medicine as suggested in the guidelines is not yet feasible. The choice of treatment strategy should depend more on rapid relief of symptoms than on prognostic factors.Trial registrationNetherlands Trial Register NTR262 (registered 7 September 2005) and NTR265 (8 September 2005).Electronic supplementary materialThe online version of this article (doi:10.1186/s13075-014-0430-3) contains supplementary material, which is available to authorized users.