This work aimed at investigating the relationship between progesterone (P₄) concentrations in milk and blood in dairy buffalo. The reliability of using milk P₄ level as an easy diagnostic tool for monitoring ovarian cycles, ovarian dysfunction and pregnancy was investigated. Twenty buffalo cows (Ten pregnant and 10 non-pregnant) were used. Four hundred sixty blood samples and another 460 milk samples were analyzed for P₄ assay using RIA technique. Different correlation coefficients between P₄ levels in milk and blood serum were computed. The P₄ profiles for the different physiological cases of the animals were diagrammed.  The results showed that P4 concentrations were consistently higher (p<0.01) in almost all milk samples matched to those of blood serum. The P₄ levels in pregnant buffalo ranged from 2.0-8.5 ng/ml in serum and from 3.1-18.6 ng/ml in milk. The P₄ levels in non-pregnant animals ranged from 0.1-8.5 ng/ml in serum and from 0.1-19.9 ng/ml in milk. The differences in P₄ levels between pregnant and non-pregnant animals, as well as between milk and serum within each group were statistically significant (p<0.01). The overall mean milk P₄ level in pregnant buffalo (11.3 ±0.2 ng/ml) was 2.3 times higher than that of blood serum (4.9±0.1 ng/ml). The milk P₄ level in non- pregnant buffalo (6.8± 0.5 ng/ml) was 2.5 times higher than that of the serum (2.7±0.2 ng/ml). Milk and serum P₄ concentrations correlated significantly (p<0.001) in both pregnant (r = 0.50) and non-pregnant (r = 0.87)buffalo. The total correlation coefficient (considering all samples from both pregnant and non-pregnant animals) was 0.83 (p<0.001). The overall mean ovarian cycle length was 20.2 ± 1.5 days with a range of 12-42 days. Half of the cycles (50%) had normal length (18-24 days). Percentages of short (< 17 days) and long (>25 days) cycles were 31.8 and 18.2%, respectively. All ovulations were confined to P₄ levels ranging from 0.1- 2.9 ng/ml in milk and from 0.1-0.4 ng/ml in serum. The P₄ peak levels of the mid ovarian cycle ranged from 6.7-17.8 ng/ml in milk and from 2.6-8.5 ng/ml in serum. It was concluded that an abrupt decline in milk P₄ level to < 2.9 ng/ml in regularly monitored buffalo could be useful indicator for the occurrence of ovulation. Confirming ovulation on the basis of a single milk sample could be misleading as the accuracy of diagnosis may be interrupted by the possibly encountered cases of static inactive ovaries or ovarian follicular cysts. The P₄ assay in repeated milk samples can provide a wider vision for accurate diagnosis of the case. It was shown that all pregnant buffalo had P₄ levels > 3.0 ng/ml in milk. In addition, all buffalo cows with milk P₄ levels of < 3.0 ng/ml were non-pregnant. That is to conclude that dairy buffalo with < 3.0 ng/ml P₄ level in milk could be diagnosed non-pregnant with high degree of accuracy. Otherwise, the use of P₄ level of > 3.0 ng/ml in milk could only be suggestive for the occurrence of pregnancy. The possibly encountered cases of ovarian luteal cysts or persistent CLs could interrupt the accuracy of diagnosis. It was concluded that milk P₄ assay could be used as an efficient mean for diagnosing non-pregnant buffalo with high degree of accuracy. Otherwise, it is tentatively suggestive for diagnosing pregnant buffalo. The   milk P₄ assay could also be an easy and useful mean for monitoring ovulation, ovarian dysfunction, embryonic mortality and pregnancy in buffalo, especially under rural conditions.