Using four sets of Bangladesh Demographic and Health Survey data collected during 1993/94, 1996/97, 1999/00, and 2004, the paper examines the trends and determinants of adolescent childbearing in Bangladesh, and identifies area-level variation in explaining differentials in adolescent first birth. Discrete-time multilevel hazard modeling is used to estimate the hazard of first birth before age 20 after controlling the effects of other individual and household factors. The results suggest that the overall probability of first birth before age 20 among Bangladeshi women remained static or even increased slightly over time. There was a significant area level variation in teenage first birth in 1993/94 and 1996/97. However, over time the effect of area is decreasing. At the individual level, women’s education, especially higher education, has the strongest effect in delaying first birth during adolescence. Age at marriage has a strong association with age at first birth: a one-year increase in age at marriage decreases the chance of teenage first birth by 10% or more. Frequent media exposure has a significant delaying effect, and the effect is more distinct in the most recent year. These findings reinforce our understanding that the government should continue its efforts to promote female education, especially higher education. Policies should also be directed to delay age at marriage, which may in turn delay the timing of first birth.