OBJECTIVES: The value of non-bone microbiological cultures in the diagnosis of osteomyelitis with sinus tract is not clear. We aimed to establish the concordance between deep sinus tract cultures and bone cultures in cases of osteomyelitis with a cutaneous fistula. METHODS: This was a non-randomized, prospective diagnostic trial at the Orthopedic Service of the University Hospital of Geneva. Each patient with osteomyelitis with a cutaneous sinus tract had four microbiological samples taken: two consecutive sinus tract cultures with bone contact at different times (samples A-1 and A-2), surgical bone biopsy performed through the sinus tract (sample B), and a surgical bone biopsy performed through an uninfected area outside the sinus tract (sample C), the latter considered as the 'gold standard'. RESULTS: One hundred and forty-one patients with 154 episodes of osteomyelitis were included in the study. When both sinus tract cultures yielded the same microorganism (86.4%), the concordance between both samples A and sample C was 96%. In the case of identical sinus tract culture infections, sensitivity was 91%, specificity 86%, and accuracy 90%. The accuracy in monomicrobial infections (50%) was higher than in polymicrobial infections (94.3% vs. 78.9%, respectively; p=0.02). CONCLUSIONS: In cases of monomicrobial osteomyelitis with sinus tract, two concordant tract cultures with bone contact accurately predict the pathogen.