The objective of the present study was to examine if luteal blood flow (LBF) monitoring could be used as an additional prognostic tool for early pregnancy diagnosis, and we particularly focused on the differences in LBF between pregnant and nonpregnant mares. Furthermore, other possible developmental differences of corpus luteum (CL) between pregnant and nonpregnant mares were evaluated. The CL (n = 119) of 27 mares were monitored once daily in B- and Power-Doppler Mode on days 1, 2, 9, 12, and 16 after ovulation (day 0 = ovulation). The data were evaluated using the MIXED Linear Model with repeated measures, and parameters were estimated by the REML method. The course of LBF, area of CL, and pixel intensity differed in nonpregnant mares on a day-to-day basis in contrast to more stable values in pregnant mares. Further, the profiles of the courses were identical until day 9, but since day 12 the differences between pregnant and nonpregnant mares started to be prominent. The LBF, pixel intensity, and level of progesterone (P4) were similar in all mares until day 16, when smaller LBF, lower pixel intensity, and lower levels of P4 were found in nonpregnant mares (P = 0,04, P = 0,02, P < 0,05, respectively). In pregnant and nonpregnant mares the LBF was weakly (r = 0,29 in both) and pixel intensity strongly (r = 0,48 and 0,59, respectively) correlated to the levels of P4. LBF was strongly correlated to the area of CL in pregnant as well as nonpregnant mares (r = 0,72 and 0,64, respectively). In accordance with the results presented in our study we can state that LBF monitoring is not a suitable tool for early pregnancy diagnosis or prognosis as the differences between pregnant and nonpregnant mares are notable – similarly to other indicators of CL status – just after the onset of luteolysis (day 16) when embryo itself is detectable.