Comparisons of eight commercial on-farm milk progesterone tests

To evaluate eight commercial on-farm milk progesterone kits, milk samples (50 ml each of foremilk and postmilk strippings) were collected during the estrous cycle from 10 cycling Holstein cows for 24 consecutive days. Relative concentrations of progesterone were classified as low or high by comparison with standard progesterone samples supplied with each kit. The concentration of progesterone in each milk sample was determined by radioimmunoassay (RIA). Accuracy of classification into low or high levels by commercial tests was determined by the percentage of similarity with RIA values using discriminant analysis. Accuracy of the eight tests ranged from 89.0 to 98.9% for low progesterone, 74.8 to 85.6% for high progesterone, and 80.3 to 87.3% for all samples (n = 238). The percentage of fat in milk or an interaction of the percentage of milkfat by day of estrous cycle influenced commercial test results for all tests except Accufirm and Calfcheck. Progesterone levels, estimated by the test-kits, were low from 1.5 +/- 0.5 to 2.8 +/- 0.9 days before estrus (X +/- SEM) and until 4.0 +/- 0.6 to 5.9 +/- 1.3 days after estrus. These data support the principle that a single low progesterone sample cannot be used to determine proper timing of insemination. All eight commercial kits can be used to determine accurately the relative concentrations of progesterone in milk samples.