[en] Three sampling procedures were compared to determine the optimal technique for isolating mycoplasmas in cattle with respiratory diseases. The prevalence of mycoplasmas isolated from these animals is also reported. In the first group, bronchoalveolar lavage (BAL) and nasal swab cultures were compared with the corresponding lung cultures from cattle necropsied for fatal respiratory diseases (n = 20). In a second group, nasal swabs were compared with corresponding BAL cultures in living animals with recurrent respiratory pathologies (n = 49). There was complete agreement between the paired BAL and lung cultures. In contrast, nasal cultures were not representative of the mycoplasmas present in the lower respiratory airways. The relative sensitivity and specificity of the nasal swab technique compared to BAL in living animals confirmed that the nasal swab cultures were not predictive of lower respiratory airway pathogens, such as Mycoplasma bovis. BAL is considered to be the best method for isolating M. bovis in cattle with respiratory diseases as it combines reliability and feasibility under field sampling conditions. In the present study, Mycoplasma dispar (43%) and M. bovis (29%) were mainly isolated in mixed infections. This confirms the need to search for mycoplasmas in routine examinations and to take them into account in therapeutic strategies for respiratory diseases in cattle.