This cross-sectional study aimed to estimate the seroprevalence of Brucella infection among sheep and goats on small-scale farms in a peri-urban area of Tajikistan and identify factors associated with seropositivity. The study population was 667 female sheep and goats >6 months of age from 21 villages in four districts surrounding the capital city, Dushanbe. Individual blood samples were collected during October and November 2012 and analysed with indirect enzyme linked immunosorbent assay (ELISA). Positive samples were confirmed with competitive ELISA. To identify factors associated with seropositivity at an individual level, a generalised linear mixed model was applied to account for clustering of individuals within villages and districts. The true individual seroprevalence was 6.7 % and ranged from 1.0 to 15.6 % between the four districts. Fourteen villages had at least one seropositive sheep or goat, resulting in apparent prevalence of 67 % at village level. The seroprevalence at individual level was significantly lower in Rudaki district (odds ratio (OR) = 0.1; 95 % confidence interval (CI) 0.03–0.4) and Varzob district (OR = 0.3; 95 % CI 0.09–0.8) than in Vahdat district. Sheep were more likely than goats to be seropositive (OR = 2.7; 95 % CI 1.3–5.5). Increasing age was significantly associated with seropositivity (OR = 1.4; 95 % CI 1.2–1.6). These results indicate high prevalence of Brucella infection among sheep and goats in the peri-urban area of the capital city in Tajikistan. Given the dense human population in such areas, this could constitute a threat to public health, besides causing significant production losses.