Abstract Background Dam-to-piglet transmission plays an important role in the epidemiology of enzootic pneumonia on farms. Although Mycoplasma hyopneumoniae (M. hyopneumoniae) infections in breeding animals are often subclinical, their control could have a positive effect on M. hyopneumoniae infection levels in fattening pigs. This study investigated the presence of M. hyopneumoniae in the breeding population of ten Belgian farrow-to-finish farms suspected by the herd veterinarian to be M. hyopneumoniae infected. Gilt vaccination against M. hyopneumoniae prior to first insemination was practiced on nine of the ten farms. At four different time points in the reproductive cycle 20 animals were sampled on each farm, namely 30–40 days of gestation, 75–85 days of gestation, 3–5 days after farrowing, and 1–3 days after weaning. In total, tracheobronchial swabs and blood samples were collected from 344 gilts and 456 sows (n = 80/farm). Swabs were analysed for the presence of M. hyopneumoniae DNA using nested PCR and M. hyopneumoniae-specific antibodies were detected in serum with a commercial ELISA. Generalized linear mixed models with farm as random factor were used to test the effect of time point in the reproductive cycle and parity on M. hyopneumoniae PCR prevalence and seroprevalence. Results M. hyopneumoniae PCR prevalence ranged between 0% and 43.8% at the farm level and the seroprevalence between 32.5% and 93.8%. Gilts were significantly more M. hyopneumoniae PCR positive than sows at the 2-4th parity (P = 0.02) and > 4th parity (P = 0.02). At 30–40 days of gestation, significantly more breeding animals were PCR positive as compared to 75–85 days of gestation (P = 0.04), 3–5 days after farrowing (P = 0.02) and 1–3 days after weaning (P = 0.02). Gilts had significantly more often M. hyopneumoniae-specific antibodies than sows (P = 0.03). Conclusions M. hyopneumoniae PCR prevalence varied a lot between farms and due to gilt vaccination the number of animals with M. hyopneumoniae-specific antibodies was high on ...