ABSTRACT A cross-sectional study was conducted to investigate seroprevalence and virus prevalence of the H9 subtype of avian influenza virus in non-vaccinated broiler farms of dense poultry-populated districts, Lahore and Sheikhupura of Punjab-Pakistan. A convenient sampling method was adopted for collection of blood (n=500) and oropharyngeal swab (n=500) samples from 25 broiler farms of each district for hemagglutination inhibition assay and RT-PCR test, respectively. Proportional estimates were calculated using R software and overall seroprevalence of H9 was estimated at 36.3% (95% CI 33.3-39), with no significant difference (p>0.05) between Lahore (37.2 %, 95% CI=31.2-39.59) and Sheikhupura (35.4%, 95% CI= 29.64-39.76). RT-PCR identified 2% (4/200) pool level viral prevalence. None of the farms from Lahore districts were RT-PCR positive for H9. Simple logistic regression followed by multivariable analysis, identified the presence of foot bath/dipping area at the entrance (OR=0.7, 95% CI=0.52-0.93) and availability of rubber shoes for visitors (OR=0.36, 95% CI 0.26-0.48) as protective factors. History of respiratory signs (OR=1.51, 95%=CI 1.12-2.04), history of sudden death in past flocks (OR=3.26, 95% CI=2.41-4.41), and birds previously infected with avian influenza virus (OR=1.33, 95% CI=1-1.76) were significant risk factors. Negligence in preventive measures at farms level was associated with the spread of H9 infection between the farms. To control future outbreaks, biosecurity and continuous monitoring of non-vaccinated flocks are suggested.