Somatic cells are present in the milk throughout lactation and consist of leukocytes and epithelial cells exfoliated from the mammary epithelium. Our objective was to determine the efficacy of using somatic cells from goat milk for dynamic studies of gene expression in the mammary gland. Over a 4-wk interval, cells were isolated from daily morning milk samples and samples taken 30 min after milking. They were characterized by direct cell counts and by flow cytometry analysis after immunostaining with antibodies directed against cytokeratin and CD45, a common leukocyte antigen. Epithelial cell counts within the morning milk ranged from 15 to 45% of total milk somatic cells. After-milking samples contained twice as many cells as did morning milk samples. The RNA was extracted from the somatic cells of both types of milk samples with equivalent efficiency (a mean of 1.2 microg RNA/mL of milk). Four mRNA variants of the alpha-S1 casein gene were detected by Northern blot analysis and the amount of each mRNA in milk cells was related to protein concentration in milk. The comparison between mRNA from the mammary gland and from congruently collected milk cells showed that relative amounts of mRNA for each milk-protein (alpha-S1 and kappa-casein and alactalbumin) were conserved. In a third experiment, daily milk cell RNA preparations were extracted to assess the effect of growth hormone (GH) on mammary gene expression; four goats were separated into two groups in order to perform a switch-back design consisting of three treatment weeks: Control, GH-Control or GH-Control-GH. In this study, treatment of goats with GH increased milk yields by 5%. Throughout the control and GH treatments, the expression of the three milk-protein genes studied were highly and significantly correlated (r = 0.949 and r = 0.958, P < 0.001 for, respectively, alpha-S1 and kappa-casein and for alpha-S1 casein and alpha-lactalbumin). During GH treatment, the three milk-protein mRNA abundances increased with the same pattern. In conclusion, the opportunity to use milk somatic cells for RNA preparation and analysis provides a significant improvement over the use of biopsy samples in assessing gene activity in the mammary gland and allows easy and repetitive sampling without damaging mammary tissue. Furthermore, we propose that this method could be used to investigate the transcriptional status of the mammary gland of an animal in relation to its genotype, nutritional and pathologic status, and under influence by hormonal factors.