Addressing potential weaknesses in an earlier investigation, the authors examined the latent structure of marital discord using 4 product indicators from the Marital Satisfaction Inventory-Revised (Snyder, 1997) in a representative sample of community couples (N = 1,020). Results from 3 taxometric procedures suggested that marital discord is taxonic, with an estimated base rate of 0.31. Cut scores for classifying members of this marital discord taxon were identified. Additional analyses with several data sets of community and clinic couples provided evidence that the marital discord taxon classification demonstrated good (a) 6-week test-retest reliability, (b) criterion validity (i.e., differences in taxon prevalence in community vs. clinic couples), and (c) construct validity (i.e., correlations with therapist ratings of relationship quality). These results suggest that discordant couples differ qualitatively and not just quantitatively from nondiscordant couples and that user friendly methods, suitable for a range of research and clinical applications, can be used to identify marital discord.