Objectives Many children with constipation who are evaluated in emergency departments (EDs) receive an abdominal radiograph (AR) despite evidence-based guidelines discouraging imaging. The objectives of this study were to identify predictors associated with obtaining an AR and to determine if ARs were associated with a longer length of stay (LOS) among children with constipation evaluated in the ED. Methods A review of billing and electronic health records was conducted in an academic pediatric ED for children ages 0 to 17 years who had a primary discharge diagnosis of constipation from July 2013 to June 2014. Logistic regression was used to identify predictors for obtaining an AR. Differences in mean LOS were analyzed using linear regression. Results In total, 326 children met inclusion criteria, and 60% of the children received an AR. In logistic regression, significant predictors included age (odds ratio [OR] = 1.1/year of age, P = 0.004), presenting with abdominal pain as chief complaint compared with constipation (OR = 4.4, P < 0.0001), and history of emesis (OR = 2.8, P = 0.001) after controlling for provider type and previous constipation medication use. In linear regression, the adjusted mean LOS for those with an AR was 163 minutes compared with 117 minutes for those without after controlling for age, provider type, and history of constipation medication use (P < 0.0001). Conclusions Abdominal radiographs were used frequently in the ED diagnosis and management of constipation, particularly in older children and those with abdominal pain and emesis. Abdominal radiographs were associated with increased LOS.