Cross-sectional prevalences and risk factors for Brucella seropositivity in goats in eastern and western Uganda were investigated. Serum was collected from 1518 goats randomly selected from 145 herds which had been identified using multistage sampling. The brucellosis card test (CT) and the Brucella melitensis tube-agglutination test (TAT) were used in parallel to detect antibodies against B. abortus and B. melitensis, respectively. Interviewer-administered questionnaires were used to collect information on goat health and management. This information was used in multivariable logistic-regression models to determine the risk factors for Brucella seropositivity in goat herds. For each analysis, a herd was considered positive if at least one goat in the herd tested positive for antibodies against Brucella and negative if none was positive. Four percent (55/1480) of the goats screened with the CT had antibodies against Brucella. The reactors were distributed in 13% (19/145) of the herds. The most-important herd-level risk factors identified were use of a hired caretaker as the primary manager of the operation compared to owner/family members (adjusted odds ratio (OR)=8.1; 95% CI 1.6, 39.7), keeping sheep in addition to goats (OR=6.0; CI 1.5, 23.7) compared to having no sheep, and free browsing (OR=4.7; 95% CI 1.0, 20.7) when compared to tethering or zero-grazing. Using the TAT, 10% (141/1446) of the goats tested positive. The positives were distributed in 43% (63/145) of the herds. Free browsing (OR=6.7; 95% CI 2.7, 16.9) when compared to tethering or zero-grazing and lack of veterinary care (OR=2.9; CI 1.3, 6.7) were the most-important factors identified in the multivariable model for B. melitensis herd seropositivity. To explore/reduce the risk of misclassification in a secondary analysis, herds were reclassified as positive if at least one goat tested positive on both tests and negative if none of the goats was positive on any of the two tests. Using this classification, 2% (30/1320; 95% CI 2, 3%) of the goats tested positive resulting in 13% (12/93) of the herds being positive. The distribution of the above risk factors by brucellosis herd-status (as defined by the second criterion) is also presented.