Knowledge about the developmental pattern of small follicles (<4 mm) will be crucial to understanding the endogenous control of folliculogenesis and to developing methods to control it for clinical purposes. This study was designed to characterize the developmental pattern of 1–3 mm follicles and to determine, retrospectively, the stage at which the future dominant follicle first attains a size advantage among follicles in the cohort. In Experiment 1, the ovaries of Hereford-cross heifers (n = 18) were examined daily by high resolution transrectal ultrasonography for one natural interovulatory interval to evaluate changes in the number of 1–3-mm follicles in relation to the wave pattern of follicles ≥4 mm. Interovulatory interval was divided into 2- and 3-wave based on number of waves exhibited by heifers. In Experiment 2, the ovaries of Hereford-cross cows (n = 9) were examined every 6 h from Day 5 to Day 13 (Day 0 = ovulation) to monitor precisely the diameter changes of individual follicles ≥1 mm during emergence of the second follicular wave. Data were analyzed by Proc. Mixed procedure for repeated measures (Littell RC et al., 2000 Stat in med 19, 1793–1819) in the Statistical Analysis System software package (SAS version 8.2 for MS Windows;; 2002 SAS Institute Inc. Cary, NC, USA). Results of Experiment 1 revealed a day effect (P < 0.05) on the number of small (1–3 mm) follicles, with a peak (P < 0.05) 1 or 2 days before wave emergence (defined as the day the dominant follicle was first detected at 4 mm), followed 3 to 4 days later by a peak (P < 0.05) in the number of large follicles (≥4 mm). There was an inverse relationship between the number of small and large follicles during Wave 1 (r = -0.66; P = 0.05) and Wave 2 (r = -0.62; P = 0.04) in 2-wave interovulatory intervals. Similarly, an inverse relation was detected between the number of small and large follicles for Wave 1 (r = -0.79; P = 0.01) and Wave 3 (r = -0.90; P < 0.01) but not for Wave 2 (r = -0.57; P = 0.14) in 3-wave interovulatory intervals. The number of 1–3-mm follicles detected in anovulatory waves did not differ (P = 0.53) between 2- versus 3-wave interovulatory intervals;; however, a difference (P < 0.05) was observed between anovulatory and ovulatory waves in 3-wave interovulatory intervals but not (P = 0.63) in 2-wave interovulatory intervals. Experiment 2 permitted the identification of the future dominant follicle at a diameter of 1 mm and its emergence at 6–12 h earlier than the largest subordinate follicle (P < 0.01). Emergence of the future dominant (r = 0.71; P = 0.05) and 1st subordinate (r = 0.78; P = 0.02) follicles was temporally associated with a wave-eliciting rise in circulating concentrations of FSH. Growth rate of the dominant and the 1st subordinate follicle differed (P < 0.01) from 2nd subordinate follicle at 84 h after their detection at 1 mm. It was concluded that small antral follicles (1–3 mm) develop in a wave-like manner, and apparent selection of the dominant follicle was manifest much earlier than previously documented.