Introduction: The Red Reflex Test (RRT) consists of visualizing the reflection that light causes in the retina when it passes through the pupil. It is a screening test for pathologies that can lead to blindness, which still have great social and economic impact on families and the government. Teaching through simulators allows students and health professionals to acquire and improve their clinical skills. Objective: The aim of the study is to develop a low-cost dummy, 3D printed and based on the Arduino platform, for the training of the red reflex technique (RRT) in newborns and evaluate its educational effectiveness. Method: A RRT dummy was presented to 7 expert judges: 6 pediatricians and 1 ophthalmologist to evaluate its applicability in medical learning. For this purpose, they used a 14-item Likert scale of 5 points. Subsequently, 40 individuals participated in a course: 33 medical students, 5 pediatric residents, a nurse and a general physician. The participants were randomly assigned to two groups: control (CG) and experimental (EG) group. Each group consisted of 20 participants. The EG went through 4 stages: 1) theoretical background; 2) handling of the simulator; 3) simulated clinical practice and; 4) evaluation in real-life patients. Whereas the CG went through the following teaching methodological processes: 1) theoretical background; 2) direct training in real-life patients and; 4) practical evaluation in a real-life patient. After each respective group intervention, both groups were assessed on their learning in 40 newborns at Santa Casa de Misericórdia do Pará Foundation in Belém, state of Pará, Brazil. Results: At the judges’ evaluation regarding the simulator positive items (design, similarity, interest, relevance, content, memorization, didactics, previous reminiscence, comprehension and application), 49.2% answered ‘strongly agree’ and 44.4%, ‘agree’. Regarding the negative questions (difficulty in understanding, information overload, abstraction, difficulty in handling and clarity of operation), 40.0% answered ‘strongly disagree’, and 57.1%, ‘disagree’. The judges were 94.9% favorable to the use of the simulator in medical education. However, comparing the two groups of students, regarding the time of exam, the results showed no statistically significant difference (p-value = 0.29). Conclusions: The dummy showed its applicability for the learning of RRT, with the advantage of being able to perform the exam without having to disturb a real-life patient.