MOST Mycoplasma species found in calf respiratory tracts are considered to be opportunistic pathogens; however, several species have been described as the primary agents for respiratory diseases (Simecka and others 1992). Mycoplasma bovis and Mycoplasma dispar have been considered the most important agents of pneumonia in cattle herds in which Mycoplasma mycoides subspecies mycoides SC is absent (Rebhun and others 2000). This short communication described a study to determine the prevalence of Mycoplasma species in the respiratory tracts of calves in Brazil. A total of 301 nasal swabs were collected from calves of both sexes, up to one year of age, from 10 different farms: seven in the State of Sao Paulo, two in Minas Gerais and one in Bahia. The samples were collected from 155 animals with respiratory disease and 146 clinically healthy animals by vigorously rubbing a swab over the surface of the nasal cavity. The nasal mucus samples were transported in 3 ml of modified Friis medium (Friis 1971) at 4°C and then incubated at 37°C for four days. The DNA was extracted from the samples by the method described by Fan and others (1995). The primers GPO-3 and MGSO (Van Kuppeveld and others 1992) were used in a PCR for the detection of Mollicutes. Samples that were positive for Mollicutes were submitted to other PCR assays using primers to detect M bovis (Gonzalez and others 1995), M mycoides subspecies mycoides SC (Dedieu and others 1994), M dispar (Marques and others 2007), and Ureaplasma diversum (Cardoso and others 2000). The PCR products were analysed on a 1·5 per cent agarose gel containing 0·5 μM/ml ethidium bromide in TAE buffer (40 mM Tris-acetate, 2 mM EDTA, pH 8·0). Table 1 shows the prevalence of Mollicutes and the four species investigated. Mollicutes were detected in 76 (52·06 per cent) of the nasal mucus samples from healthy animals and in 141 (90·96 per cent) of the samples from animals with respiratory disease. Although M dispar and U diversum were detected in both sick and healthy animals, M bovis was detected only in calves with respiratory disease. M mycoides subspecies mycoides SC was not detected in any of the samples. M dispar and U diversum were detected more frequently in samples from animals with respiratory disease than healthy animals. Of the 301 nasal mucus samples, 102 (33·88 per cent) were positive for at least one of the species tested for. The frequency of Mollicute co-infections in healthy and sick animals is shown in Table 2. M mycoides subspecies mycoides SC is the classical primary agent of contagious bovine pleuropneumonia. According to the World Organisation for Animal Health (OIE), healthy cattle herds must be free of this microorganism. This species was not detected in the present study; however, the authors recommended that a wide-ranging investigation using PCR should be carried out in Brazil. M bovis has previously been isolated in healthy animals, and was originally described as a natural microorganism of the bovine respiratory system (Knudtson and others 1986, Ter Laak and others 1992b). Currently, M bovis is the second most important mycoplasma in bovine respiratory diseases (Gevaert 2006). In the present study, M bovis was detected only in sick animals, either as a single agent or in association with M dispar or U diversum. The results obtained are in agreement with other studies that have mentioned M bovis as a species associated with respiratory disease outbreaks in cattle herds (Bashiruddin and others 2001, Byrne and others 2001). M dispar was detected more frequently in sick calves than in healthy animals. Ter Laak and others (1992a) reported M dispar to be the main species isolated from lung samples from sick calves (49 per cent of the samples). In contrast, Ter Laak and others (1992b) showed a frequency of 40 per cent of M dispar detection in healthy animals. In the present study, U diversum was detected by PCR in 12 (8·22 per cent) nasal mucus samples from healthy calves, and in 49 (31·6 per cent) samples from sick animals. Ter Laak and others (1992b) also detected U diversum in healthy animals as well; however, the frequency of detection was higher in sick animals. In the present study, healthy animals that were PCRpositive for M dispar and U diversum were in close contact with sick animals infected with these species, suggesting that M dispar and U diversum may be important pathogens in the development of respiratory disease. The detection of Mollicutes in nasal mucus from both sick and healthy calves may indicate the opportunistic role of these bacteria. The host-parasite relationship of mycoplasmas is a complex subject, especially in calves, in which the disease process in the respiratory tract is not well understood. Co-infections with mycoplasmas in the respiratory tracts of calves are common (Thomas and others 2002, Levisohn and others 2004). In the present study, three mycoplasma species were detected together in four samples, but an association between U diversum and M dispar was most common. Brazil is one of the most important cattle meat producers in the world. Consequently, it is important to monitor disease occurrence among Brazilian herds, including mycoplasma infections. The PCR results for identifying mycoplasmas in TABLE 2: Frequency of co-infections with more than one Mycoplasma species in the respiratory tracts of healthy calves and calves with respiratory disease