The objectives of this study were to relate concentrations of plasma (PUN) and milk (MUN) urea nitrogen to pregnancy rate in dairy cows and compare various methods of analysis and preparation of milk for measuring MUN. In two experiments, blood or milk samples were collected on the day of AI from Holstein cows (n = 160 and n = 155, respectively). Three methods of MUN analysis were compared. Two laboratory chemical procedures yielded similar results, whereas a quick dipstick method overestimated chemical analyses. Before and after milking strip samples had MUN concentrations equivalent to those in composite milk. Concentrations of PUN or MUN greater than 19 mg/dL were associated with decreased (P < .02) pregnancy rates (18 and 21 percentage point reduction in the two experiments). In two subset groups of cows (n = 51 and n = 23, respectively), plasma progesterone or MUN concentrations were monitored during the 5-d period after AI. Plasma progesterone concentrations increased similarly during the period for cows divided into low vs high PUN but were greater in pregnant than in nonpregnant cows on d 4 and 5 (P < .04). The MUN concentrations showed low within-cow variation (CV = 8%) but were lower in pregnant cows and had a decreasing trend over time compared with nonpregnant cows (P < .05). Based on this study, plasma and milk will yield similar results for monitoring urea nitrogen in dairy cows; PUN and MUN concentrations > 19 mg/dL were associated with approximately a 20 percentage point decrease in pregnancy rate after AI in lactating dairy cattle.