Two hundred and fifty-one cows fresh at least 37 days with normal reproductive tracts and palpable ovarian structures which clinicians believed to be corpora lutea and which were presumed to be producing progesterone were treated with 500 micrograms of cloprostenol. The cows were stratified into three groups based on milk progesterone concentrations in whole milk samples taken at the time of examination. The low group had milk progesterone concentrations less than or equal to 1 ng/ml, the intermediate group had milk progesterone concentrations between 1 and 3 ng/ml, and the high group had milk progesterone concentrations greater than or equal to 3 ng/ml. The proportion of cows in each group inseminated within 5 days of treatment and the fertility at that breeding were compared. There were no significant differences among the groups with respect to the proportion of cows that came into estrus within 5 days. However, the conception rate of the high group (54%) was significantly greater than that of the intermediate group (26%). To study the relationship between practitioner experience and diagnostic accuracy the proportion of cows with low, intermediate and high concentrations of progesterone selected by clinicians with more than 3 years experience was compared to that for clinicians with 3 years of experience or less. Less experienced clinicians were significantly more likely to make a false positive diagnosis of functional luteal tissue in cows with palpable ovarian structures than were more experienced clinicians.