The maned wolf (Chrysocyon brachyurus), a neotropical canid, lives in habitats severely compromised by agricultural development. Therefore, maintaining a healthy and viable zoo population is crucial, because it serves as hedge against extinction. However, the North American ex situ population of maned wolves is far from self-sustaining, mainly due to low pregnancy success and high neonatal mortality. Currently, there is no non-invasive way to determine pregnancy in this species. Therefore, the goal is to determine whether urinary relaxin can be used as an indicator of pregnancy in this species. Objectives were to (1) determine the relationship between serum and urinary relaxin in canids, using pregnant domestic bitches as a model, and (2) evaluate the patterns of urinary relaxin in pregnant v. non-pregnant female maned wolves. Serum and urine samples were obtained weekly from six pregnant domestic bitches starting from the day of natural breeding until 3 weeks before the estimated whelping dates. For the maned wolf, urine and feces were obtained weekly during a breeding season (October to February) from six females, three of which were pregnant and gave birth. Concentrations of serum and urinary relaxin were measured using the double antibody canine relaxin radioimmunoassay (Steinetz et al. 1996 Biol. Reprod. 54, 1252–1260). Fecal progestagen were determined using an enzyme immunoassay (Songsasen et al. 2006 Theriogenology 66, 1743–1750). The relationship between serum and urinary relaxin concentrations was determined using the Pearson Product Moment Correlation Analysis. The relationship between urinary relaxin concentrations and days before whelping (for pregnant females) or time after the initial rise of progestagen for each individual were determined using a linear regression analysis. Serial dilutions of pooled dog urine produced displacement curves parallel to those of the appropriate synthetic canine relaxin standards, while wolf samples showed limited parallelism with low concentration of immunoactive antigen. Concentrations of urinary relaxin were significantly correlated with the serum counterpart in pregnant dogs (P < 0.01, r2 0.58). In pregnant maned wolves, relaxin level increased two- to six-fold as gestation progressed (P < 0.05, r2 ranged from 0.15 to 0.56). In contrast, urinary relaxin remained at the baseline in non-pregnant females throughout the diestrus period (P > 0.05). In conclusion, our findings suggest that urinary relaxin mirrors serum relaxin during pregnancy in domestic bitches and may have potential as a tool to diagnose pregnancy in the maned wolf. Further studies are required to improve the specificity and sensitivity of this non-invasive assay.