RESUMEN Objetivo: Estimar la asociación entre la adicción a internet (AI) y el rendimiento académico de los estudiantes de Odontología de la Universidad de Cartagena. Material y métodos: Estudio de corte transversal en 402 estudiantes, seleccionados de modo no probabilístico, que respondieron a un cuestionario anónimo autoaplicado que incluía variables sociodemográficas, rendimiento académico (promedio acumulado del último semestre), presencia de AI (test de Young) y variables relacionadas con la AI en función del rendimiento académico. Los datos se analizaron a partir de proporciones, relaciones entre variables con test de la X 2, y la asociación se obtuvo por razones de disparidad (OR) a través de regresión logística nominal. Resultados: El 75,3% de los estudiantes mostraban AI; el 24,63% utilizaba internet mucho menos que la población promedio; el 73,13% mostraba una AI leve; el 2,24%, una AI moderada y no hubo casos de AI grave; el 5,2% tenía bajo rendimiento académico. En el análisis multivariable, el modelo que mejor explica la AI en relación con el rendimiento académico fue: estudiar en semestres inferiores (OR = 0,54; IC95%, 0,32-0,91), estudiar en lugar distinto de la casa (OR = 3,38; IC95%, 1,71-6,68), usar elemento no portátil para estudiar (OR = 0,41; IC95%, 0,19-0,89), chatear por celular (OR = 2,43; IC95%, 1,45-3,06) y demorar más de 18 min (OR = 3,20; IC95%, 1,71-5,99) mientras se estudia. Conclusiones: El rendimiento académico no se asocia con la AI. Sin embargo, estudiar en semestres inferiores, en un lugar distinto de la casa, emplear elementos no portátiles para estudiar e invertir más de 18 min en contestar el celular y chatear mientras se estudia son covariables estadísticamente asociadas con la AI. ABSTRACT Objective: To determine the association between Internet addiction (IA) and academic performance in dental students at the University of Cartagena. Material and methods: A cross-sectional study was conducted in 402 students included through non-probabilistic sampling who answered an anonymous and self-reporting questionnaire that included socio-demographic variables, academic performance (last semester overall grade), presence of IA (Young's Test) and covariates related to IA based on academic performance. Data were analysed by means of proportions, relationships between variables with the X 2 test and strength of association was estimated with odds ratios (OR) using nominal logistic regression. Results: Approximately 24.63% of the students used the Internet much less than the average population, but 75.3% had IA; 73.13% of cases were considered mild and 2.24% moderate. There were no severe cases. Around 5.2% had poor academic performance. In multivariate analysis, the model that best explained IA in relation to academic performance was: studying in lower-level courses (OR=0.54; 95% CI, 0.32-0.91); studying in a different places of the house (OR=3.38; 95% CI, 1.71-6.68); not using laptop for studying (OR=0.41; 95% CI, 0.19-0.89), chatting on mobile phone (OR=2.43; 95% CI, 1.45-3.06); and spending more than 18 minutes on mobile phone while studying (OR=3.20; 95% CI, 1.71-5.99). Conclusions: Academic performance was not associated with AI. However, studying in lower-level courses, in a different place of the house, not using laptop to study, and spending more than 18 minutes answering their mobile phone and chatting on mobile phone while studying were covariates statistically associated with IA.