The objective of this study was to determine if milk samples, as they are routinely collected by Ontario Dairy Herd Improvement, would yield accurate milk urea results with an infrared assay. This investigation involved analytic validation of the infrared assay and assessment of the effect of DHI routine sample acquisition factors on milk urea results. Analytic validation of an automated milk urea assay was performed by assessing the relative accuracy and precision of milk urea results produced by the Fossomatic 4000 Milk Analyzer, an infrared method of analysis, compared with the Eurochem test, an accepted reference method. Results indicated that, when interpreted at the group level, milk urea results between the infrared method and the reference test were in good agreement. The two tests shared a similar and high level of precision. Milk urea concentrations obtained from composite (metered) milk samples, and not quarter stripping samples, were most representative of concurrent serum urea concentrations. The addition of bronopol preservative did not result in a numerically important change in milk urea concentrations. Storage of preserved metered milk samples for up to 4 d at either room temperature or by refrigeration, or for up to 3 d by freezing, did not result in changes in milk urea concentrations. We concluded that milk samples, as they are routinely collected and handled by DHI, are suitable for measurement of milk urea concentrations with the infrared method of analysis if data are interpreted at the group level.