Our objective was to estimate the prevalence and evaluate factors associated with smoking among high school adolescents in Karachi, Pakistan. A school-based, cross-sectional survey was conducted in three towns in Karachi, namely, Gadap Town, Bin-Qasim Town and Malir Town, from January through May 2003. Two-stage cluster sampling stratified by school type was employed to select schools and students. We recruited and interviewed 772 male students regarding socio-demographic factors, smoking history of students, their families/friends, number of siblings, and place of residence. The prevalence of smoking (30 days) among adolescents was 13.7%. Final multiple logistic regression analysis showed that after adjusting for age, ethnicity, and place of residence, being a student at a government school (OR=1.6; 95% Cl: 1.0-2.7), parental smoking (OR = 1.7; 95% Cl: 1.1-2.8), uncle smoking (OR = 1.7; 95% Cl: 1.2-2.8), peer smoking (OR = 6.2; 95% Cl: 3.9-9.9), and spending leisure time outside home (OR = 3.9; 95% Cl 1.2-13.2) were significantly associated with adolescent smoking.