Objective
To estimate the adult prevalence of HIV among the adult population in Zambia and determine whether demographic characteristics were associated with being HIV positive.

Methods
A cross sectional population based survey to asses HIV status among participants aged 15 years and above in a national tuberculosis prevalence survey. Counselling was offered to participants who tested for HIV. The prevalence was estimated using a logistic regression model. Univariate and multivariate associations of social demographic characteristics with HIV were determined.

Results
Of the 46,099 individuals who were eligible to participate in the survey, 44,761 (97.1 %) underwent pre-test counselling for HIV; out of which 30,605 (68.4 %) consented to be tested and 30, 584 (99.9 %) were tested. HIV prevalence was estimated to be 6.6 % (95 % CI 5.8–7.4); with females having a higher prevalence than males 7.7 % (95 % CI 6.8–8.7) versus 5.2 % (95 % CI 4.4–5.9). HIV prevalence was higher among urban (9.8 %; 95 % CI 8.8–10.7) than rural residents (5.0 %; 95 % CI 4.3–5.8). The risk of HIV was double among urban dwellers than among their rural counterparts. Being divorced or widowed was associated with a threefold higher risk of being HIV positive than being never married. The risk of being HIV positive was four times higher among those with tuberculosis than those without tuberculosis.

Conclusions
HIV prevalence was lower than previously estimated in the country. The burden of HIV showed sociodemographic disparities signifying a need to target key populations or epidemic drivers. Mobile testing for HIV on a national scale in the context of TB prevalence surveys could be explored further in other settings.