Abstract The controversial data about antral follicle count (AFC) may be partially explained by the different criteria used to determine what is high, intermediate and low AFC. This study evaluated different classification methods for AFC groups, relating them to the conception rate, dominant follicle size and body condition score (BCS) in cows submitted to timed artificial insemination (TAI). Nelore cows (Bos indicus; n = 935), received a reproductive program consisting of TAI and natural breeding. Conception rate, BCS and dominant follicle size during TAI were evaluated by three AFC methodologies: i) mean and standard deviation: low (≤ 15 follicles); intermediate (≥ 16 to ≤ 44 follicles) or high (≥ 45 follicles); ii) quartiles: low (≤ 15 follicles); intermediate (≥ 16 to ≤ 39 follicles), or high (≥ 40 follicles); and iii) AFC score: I (low; ≤ 15 follicles); II (intermediate; ≥ 16 to ≤ 30 follicles); III (high; ≥ 31 to ≤ 44 follicles) or IV (very high; ≥ 45 follicles). Data were analyzed by a GLIMMIX and Tukey test or binary logistic regression model (P ≤ 0.05). The conception rate to TAI was influenced (P < 0.05) by AFC in the three methods classification, being the highest conception rate observed in the low AFC group regardless of method utilized: Mean (low 61.73%a, intermediate 54.02%ab and high 49.48%b), Quartiles (low 61.73%a, intermediate 53.59%ab and 51.46%b) and Score (I 61.73%a, II 54.80%ab, III 53.23%ab and IV 49.48%b). There were variations (P < 0.05) in the conception rate within the 2.50 to 2.75 BCS range for all AFC classification methods, with the low AFC females presenting the best results, regardless of the method used. Also, females with low AFC showed larger (P < 0.05) diameters of dominant follicles at the TAI regardless of method. The different methodologies used (Mean, Quartile and Score) to AFC classification showed a consistency between the main findings, and we believe that this standardization will facilitate the interpretation of data involving AFC.