Matrix metalloproteinases (MMP) and tissue inhibitors of MMP (TIMP) regulate tissue remodeling events necessary for ovulation. Thus, changes in MMP and TIMP expression and protein enzyme activity were examined in vivo and in vitro during follicular development and atresia in the horse. Equine granulosa cells and follicular fluid from medium (15 to 29 mm) healthy and atretic follicles and from large (>30 mm) healthy and preovulatory follicles were collected by transvaginal aspiration. The cells were either snap-frozen (in vivo study) or cultured for 48 h (in vitro study) to determine gene expression and protein enzyme activity of MMP-2 and MMP-9 and TIMP-1 and TIMP-2. Concentrations of progesterone and estradiol were determined by RIA in follicular fluid and conditioned media and were used along with follicle dynamics to classify follicles. In vivo, expression of MMP-2 and TIMP-2 was increased (P < 0.05) in large-preovulatory follicles, whereas TIMP-1 was decreased. The ratio of MMP-2:TIMP-2 expression was decreased (P < 0.05) in medium-healthy and large-preovulatory follicles, whereas the MMP-9:TIMP-1 ratio was increased only in large-preovulatory follicles compared with large-healthy follicles. Estradiol was greatest (P < 0.05) in the fluid of large-healthy and large-preovulatory follicles. However, medium-atretic follicles were associated with the least estradiol concentrations, both in vivo and in vitro. Progesterone concentrations were greatest (P < 0.05) in large-preovulatory follicles both in vivo and in vitro. In healthy follicles in vivo, the diameter was correlated with estradiol concentration, the estradiol:progesterone ratio, MMP-9 and TIMP-1 expression, and MMP-2 and MMP-9 protein activity. In contrast to in vivo studies, the ratio of MMP-9:TIMP-1 expression was increased (P < 0.05) in medium-healthy follicles; TIMP-2 expression decreased in large-preovulatory follicles in vitro. In addition, MMP-9 protein activity was decreased (P < 0.05) in the media samples of cells from large-healthy follicles compared with those from medium-healthy follicles. These results indicate that changes in MMP-2 and MMP-9 activities may be essential to the tissue reorganization necessary for ovulation in the equine ovary.