With the aim of evaluating the effect of freezing goat milk samples on recovery of intramammary pathogens, 1200 milk samples from udder halves with subclinical intramammary infection were studied. Samples (20 ml) were frozen at -20 and at -80 degrees C. Thawing was carried out at room temperature at 7, 14, 21, 28, 58, 118, 178, 236 and 730 days after collection and bacteriological analyses were carried out to determine the number of colony forming units/ml (CFU/ml). Mixed model statistical analysis showed that bacterial group, temperature of storage, interaction of bacterial group and temperature of storage and the interaction of bacterial group, time and temperature of storage were statistically significant effects. For coagulase negative staphylococci (CNS), least squares means of log CFU/ml recovered at -20 and -80 degrees C were not different. Nevertheless, for Gram negative bacilli (GNB) a significant decrease was detected in samples frozen at -20 vs. -80 degrees C. At both temperatures and at different times of storage, significant increases were detected between log CFU/ml of CNS and values on day zero. At -20 degrees C, a significant decrease in GNB recovery was detected between freezing days zero and 730. This difference was not detected when goat milk samples infected by GNB were frozen at -80 degrees C. The results show that frozen milk samples can be useful in goat subclinical mastitis control programs.