When operational modal analysis (OMA) is conducted on operating machines or structureswithout knowing excitations and perturbations neither the number of concerned frequencies,one problem raises on the validity of modal parameters identification (especially dampingratios) and on the precision of results obtained. This paper presents an OMA methodallowing for the evaluation of uncertainties of modal parameters (frequencies, dampingratios and mode shapes). The method is based on the vector autoregressive model (VAR) formultiple numbers of measured channels. It is seen that the uncertainty of modal parametersdecreases with higher model orders; the calculation of uncertainty allows also for theconstruction of an objective criterion for the selection of computing model order based ona threshold of confidence interval. The derivation confirms also that the identificationof natural frequencies deals with a smaller uncertainty compared with the damping ratioestimation, and hence it can be conducted with a lower computing model order. Numericalsimulations and experiments on a steel plate show the feasibility and the effectiveness ofthe developed OMA method.