The purpose of this study was to evaluate the efficacy and accuracy of mandated reporters to identify child abuse in children presenting with fractures. An Institutional Review Board approved survey-based study between January 2017 and December 2017 was conducted at a tertiary care academic medical center. 10 cases were combined to create one survey. Each case had information on presentation history, radiographic data, and social history. This study assesses the ability of 13 orthopedic residents and 11 medical students to diagnose child abuse. Participants had the option to explain their reasoning for a given case. To evaluate decision-making reasoning, we split responses into three cohorts, encompassing objective evidence, subjective evidence, or social evidence. Twenty-four participants completed the survey; 203 out of 240 (85%) included the rationale for the diagnosis of child abuse. The observed diagnostic odds ratio was 0.83 for medical students, 0.93 for junior residents, and 0.96 for senior residents. There was no statistically significant difference in diagnosing child abuse between a participant’s level of experience, age, or whether participants had their own children. Participants who used more than one source of evidence were significantly more likely to make the correct diagnosis (P = 0.013). Participant decisions were no more accurate than a coin toss. The use of several data sources led to increased diagnostic accuracy. There is low accuracy in correctly diagnosing child abuse in our cohort of mandated reporters. Participants who highlighted using several sources of evidence were more likely to diagnose child abuse accurately.