Early detection of pregnancy could enhance the reproductive efficiency of donkeys, as the species has a wide range for the first detection of the embryonic vesicle (EV) via B-mode ultrasonography. This study aimed to assess the uterine blood flow of jennies as an early indirect pregnancy diagnostic test. We hypothesized that uterine blood flow is increased, thus, serving as an early pregnancy diagnostic tool. Jennies (n=52) were bred with fresh semen from a fertile jack, and nine jennies were not bred and served as controls; ovulation (Day-0) was confirmed at 12h intervals for all jennies. Power-doppler and B-mode ultrasonography were performed on Days 7 to 10 to assess the uterine horn blood flow and identification of EV, respectively; the latter was repeated on Day 15. Uterine horn blood flow was evaluated using spot meter techniques, measuring mean pixel intensity. Progesterone concentrations were determined with RIA on Days 7 to 10. Data were assessed for normality with Shapiro- Wilk's test and mixed models. Receiving operating curve was used to determine the cut off value for uterine blood flow. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for blood flow on Days 7 to 10 were calculated to identify pregnant jennies and confirmed on Day 15. Significance was set as P≤0.05 and tendency as 0.05<P<0.1. Of 52 bred jennies, 29 (53%) were confirmed pregnant on Day 15. B-modeultrasonography identified 17.2% (5/29) of the EVs on Day 9 and 48.3% (14/29) on Day 10; no EVs were identified on Days 7 to 8. Pregnant jennies (n=29) had greater uterine blood flow in comparison with non-pregnant jennies (144.8±1.5 vs. 139.2±1.3 pixels/unit, n=32) (P=0.007), with an SD of the difference of 6.3 pixels/unit (p=0.008). There was no effect of time (P=0.48) or group x time interaction between (P=0.70) for uterine blood flow. Progesterone concentrations did not change within Days 7 to 10 (P=0.81), and there was no group x time interaction (P=0.49). Progesterone tended to be greater in pregnant than non-pregnant jennies (P=0.07). There was no effect of time (P=0.79), group (P=0.77), or group x time interaction (P=0.92) for uterine horns. Sensitivity from Days-7-to-10 were 86%, 82%, 60%, and 62%, respectively. In contrast, specificity for Days 7 to 10 was 80%, 88%, 78%, and 75%, respectively. The PPV and NPV for Days 7 to 10 were 81% and 85%; 75% and 20%; 35% and 33%; and 30% and 27%; respectively. In conclusion, this study demonstrated the efficacy of power-doppler-ultrasonography as a sensitive and specific tool for early indirect detection of pregnancy in donkeys, whereas standard B-mode first detected a small percentage of pregnancies on Days 9 to 10. Power-doppler could facilitate the breeding management of donkeys, such as timing the induction of luteolysis in non-pregnant jennies at Days 7 to 10 post-ovulation.