Objective: To determine the feasibility and effect size of an emergency department (ED)- based intervention using text messaging (TM) to improve contraception initiation among underserved adolescent females at high risk of pregnancy. Methods: This was a prospective pilot randomized controlled study. Eligible females were 14–19 years old, sexually active, presenting for reproductive health complaints and at risk for pregnancy, defined as nonuse of effective contraception. We randomized participants to standard referral (SR) or the intervention, 3 months of TMs that included reproductive health information and consistent promotion to go to family planning. The timing, frequency, and content of messages were based on qualitative interviews and a modified Health Beliefs Model. The primary outcome was contraception initiation (CI) though phone call follow-up and electronic health record review at 3 months. Results: We enrolled 100 participants; lost to follow-up was 12%. CI was 6/42(14%) in the TM group and 13/45(27%) in the SR group (absolute risk difference -0.146; 95%CI -0.309, 0.03). Using the per protocol analysis, CI was 6/36 (17%) in the TM group and 13/45 (29%) in the SR group (absolute risk difference -0.122; 95%CI -0.292, 0.067). Age of the patient modified the effect of the intervention, with younger girls who received TM more likely to initiate contraception than younger girls in the SR group (p=0.02). Conclusion: This feasible ED-based TM intervention did not affect contraception initiation rates but did show promise among the younger population. Future ED-based pregnancy prevention interventions might consider targeting younger adolescents and designing more complex interactive messaging interventions.