We conducted this cross-sectional study to investigate risk factors of Brucella seropositivity in cattle herds reared in livestock-wildlife interface areas of Blue Lagoon and Lochinvar National Parks in Zambia between August 2003 and September 2004. Sera were collected from cattle aged > or =2 years from 124 herds. Data on husbandry practices, grazing strategies, and herd structure (sex and age composition) were also collected. Sera were screened for anti-Brucella antibodies using the Rose Bengal test (RBT) as a presumptive test and a competitive-ELISA (c-ELISA) as a confirmatory test. A herd was classified as Brucella seropositive if at least one animal tested positive on both RBT and c-ELISA in series testing. Risk factors for herd-level brucellosis seropositivity were tested using multivariable logistic regression; risk factors for increases in the within-herd counts of seropositive cattle were analyzed using the negative binomial regression model with the number of seropositive animals as outcome and total number of cattle tested in a herd as the population at risk (exposure). Of the 110 herds tested, 68 (62; 95% CI: 53, 71% after adjusting for clustering by area) tested seropositive for exposure to Brucella spp. The final logistic-regression model identified geographical area, with Lochinvar (OR=3.4; CI: 0.97, 12) and Kazungula (OR=4.3; CI: 0.91, 20) recording higher odds of Brucella infections compared to Blue Lagoon. Herds coming in contact with wildlife had higher odds compared to those without contact (OR=3.4; CI: 1, 11). Similarly, the odds of Brucella infection were progressively higher in the larger herd categories (26-40 cattle, OR=2.6; CI: 0.70, 10; 41-82 cattle, OR=4.9; CI: 0.93, 26; >82 cattle, OR=9.4; CI: 1.7-51) compared to the smallest herd category (10-25). The negative binomial regression model identified geographical area, contact with wildlife, and herd size as having significant effect on counts of seropositive cattle in a herd.