Current research on follicular dynamics is focussing more on primordial and preantral follicular populations. Whereas antral follicular development can be studied by ultrasound visualization, preantral follicles can be visualized only on stained ovarian tissue sections obtained through laparotomy or following slaughter, which does not allow dynamic follicular follow-up studies. To our knowledge, only one report has previously been published on a technique for transvaginal corpus luteum biopsy (Kot K et al. 1999 Theriogenology 52, 987–993). Therefore, a new method to repeatedly take ovarian biopsies in living donor cattle through ultrasound-guided transvaginal puncture of the ovarian stroma was developed and tested. Healthy cows with a normal reproductive tract upon rectal examination were prepared as for transvaginal ovum pick-up (OPU) (Bols PEJ et al. 1995 Theriogenology 43, 677–687). The ovary was visualized using a mechanical multiple angle 5-MHz ultrasound transducer (Pie Medical, the Netherlands), fitted next to a needle guidance system in an OPU handle. A specially designed needle carrier with a 14-gauge disposable needle (length 38 mm, outer diameter 2.1 mm, Kruuse, Japan), acting as a trocar, was inserted through the vaginal wall. A 60-cm-long disposable biopsy needle with an inner stylet having a 20-mm-long specimen notch, a cutting cannula (diameter 1.3 mm), and an automated spring-loaded handle with trigger (Quick-Core^® Biopsy Needle, Cook, Denmark) was passed through the disposable needle into the ovarian stroma, under ultrasound control. Prior to insertion, the plunger of the biopsy needle was pulled back, with the needle spring-locked in the ready position. The biopsy needlepoint was positioned closely under the surface of the ovary with the stylet retracted and the specimen notch covered by the cutting cannula. With the needle in position, the stylet was advanced to expose the specimen notch within the ovarian stroma. Then, the cutting cannula was fired and a tissue sample was captured. Repeated biopsies were taken in four cows two times a week, over a two-week period, without echographically detectable detrimental effects of the reproductive tract. A tissue specimen was recovered in 25 of 30 attempts, although the dimensions of some of the biopsies were rather small. Biopsies from both the ovarian stroma and the corpus luteum (CL) were taken. They were fixed with 4% paraformaldehyde for 2 h at room temperature. Following overnight incubation in a 15% sucrose solution for cryoprotection, 17-μm cryostat sections were made and stained with hematoxylin. Study of the sections revealed small populations of primordial follicles clearly distinguishable from the surrounding ovarian stroma. Distinct groups of luteinized cells could be seen in CL biopsies. Further studies are necessary to standardize the procedure and to evaluate the effects of biopsy-taking on fertility.