It is now well established that milk progesterone analysis can provide a valuable tool in reproductive management of the dairy cow. With the advent of relatively cheap cow side milk progesterone tests, the principle block to widespread monitoring is the time and effort involved in the actual collection and analysis of the milk samples. This has led to a research drive to develop on line milk progesterone monitoring systems whereby progesterone is routinely monitored at normal milking. Near-infrared spectroscopy (NIRS) is widely used in the agricultural industry for analysing cereals and forages and is used for on-line testing in the food and pharmaceutical industries. Vibrations of chemical bonds between carbon, oxygen, hydrogen and nitrogen atoms affect the amount of near-infrared light absorbed at individual wavelengths, so different molecules show unique patterns in the spectrum. These patterns are related to determined concentrations of chemicals in calibration samples to develop equations that predict concentrations in unknown samples. This project set out to assess the use of this technique to measure milk progesterone.