Urinary nitrogen (N) excretion (UN) as a proportion of N intake (NI; UN/NI) is a major determinant of N excretion from ruminants and could be predicted from the N isotopic discrimination occurring between dietary and animal proteins (Δ 15 N). This study investigated the usefulness of Δ 15 N and other plasma biomarkers to reflect changes in UN/NI from sheep offered different levels of dietary urea. Eighteen Merino rams (age, 1–2 years; live weight, 41 ± 3 kg) were allocated to three dietary N treatments for a N balance study. Treatments were control (C), control + 0.5% urea (C+0.5%), and control + 1.2% urea (C+1.2%) and designed to provide maintenance, maintenance plus an additional 15%, and maintenance plus an additional 33% NI, respectively. The urea effect term was used for one-way ANOVA and regression analysis. As NI increased, the UN and retained N (RN) increased linearly ( p < 0.001), but UN/NI only increased in treatment C+1.2% compared with C ( p < 0.05). Plasma Δ 15 N was positively and significantly correlated with UN and UN/NI ( r = 0.52, p = 0.028; and r = 0.68, p = 0.002, respectively) and increased linearly ( p < 0.001) with the highest values observed in C+1.2%. Urine δ 15 N changed linearly between C and C+1.2%, but plasma δ 15 N increased quadratically ( p < 0.05). Plasma urea N increased in a linear way across dietary urea levels ( p < 0.001). The N isotopic difference between plasma and urine (plasma δ 15 N–urine δ 15 N) of C did not vary from either of the other treatments; however, it differed between C+0.5% and C+1.2% ( p < 0.05). The study confirmed the potential usefulness of plasma Δ 15 N to estimate UN/NI from sheep. Moreover, plasma δ 15 N–urine δ 15 N can be proposed as a new biomarker of N excretion from small ruminants. These approaches, however, need to be tested in various study conditions.