Six cross-bred mares (Mangalarga, Quarter Horse, Apaloosa, Criollo) were used to (1) quantify the serum progesterone (P4) levels during the oestrous cycle of mares using chemiluminescence (CLIA), (2) compare the P4 profile quantified using CLIA and the follicular dynamics data obtained using ovarian ultrasound (US), and (3) validate the CLIA methodology with an enzyme-linked immunosorbent assay (ELISA). The horses were between 10 and 18 years old, body weight ranging from 350 to 490 kg, and with a body condition score of 3.5. We conducted the study starting from the day of ovulation until the next ovulation was detected using US examinations. Blood samples were collected at intervals of two days and stored at -20 °C for the determination of P4 using ELISA and CLIA. Concomitantly, the ovaries were scanned daily using US to monitor the follicular dynamics. The length of the oestrous cycle (interval ovulation to ovulation) varied between 22 and 27 days. The results of the P4 concentration were analysed using Pearson's χ² and Kolmogorov-Smirnov tests. The P4 concentration in the follicular phase ranged from 0.38 to 1.91 ng/mL (CLIA) and 0.81 to 3.26 ng/mL (ELISA), and the concentrations in the luteal phase were 8.69 to 29.05 ng/mL and 11.84 to 25.16 ng/mL, respectively. The P4 concentration ≥ 5.48 ng/mL was indicative of the luteal phase (days 3 to 15 of the cycle) and < 5.47 marked the follicular phase of the oestrous cycle (day 16 until the next ovulation). The dimensions (diameter) of the largest follicle and corpus luteum varied between 35.0 and 47.5 mm and 14.2 and 43.7 mm, respectively. A follicular wave was observed during the oestrous cycle of the animals. The follicular and luteal phases of the oestrous cycle were established based on P4 concentrations after the analysis using the χ² and Kolmogorov-Smirnov adherence tests. The Pearson's χ² test results indicated a p-value < 0.05, presenting different frequencies and numerical values between ELISA and CLIA; however, the curves of both the analyses depicted the same trend. Therefore, after adjustment, the Kolmogorov-Smirnov adherence test (KS at 95 % probability) was performed where the maximum difference between the observed and estimated frequencies was compared and divided by the number of observations. The values proved to be adherent, with the same tendency in the resulting curve. The CLIA proved to be an efficient method for determining the serum P4 in mares, as it determined the P4 profile, showing adherence to the ovarian findings verified using US examinations. Therefore, CLIA can be used as an alternative method because it has been proved to be an efficient, low-cost and rapid method.