RESUMEN Objetivo: Describir los factores de riesgo asociados con mala respuesta a fenobarbital como primera linea de tratamiento en crisis neonatales en Fundacion Hospital de la Misericordia. Metodos: Estudio de casos y controles. Fueron incluidos pacientes menores de 44 semanas de edad concepcional con diagnostico de crisis neonatales tratados con fenobarbital. Controles (n=35) fueron los pacientes con adecuada respuesta a fenobarbital. Casos (n=20) fueron aquellos sin respuesta a fenobarbital y que requirieron un segundo anticonvulsivante. Resultados: Se identificaron como factores de riesgo: Mas de un tipo de crisis (OR11,625 IC95%2,946–45,877;p<0,001), latencia de inicio de fenobarbital mayor de 12 horas (OR9,917 IC95%2,205–44,607;p=0,001), presencia de crisis sutiles (OR5,167 IC95%1,309–20,39;p=0,014) o tonicas (OR3,271 IC95%1,042–10,278;p=0,039) y crisis mayores a 5 minutos de duracion (OR2,824 IC95%1,927–4,137;p=0,026). Los antecedentes perinatales que se identifi caron como factores de riesgo fueron: Apgar bajo al minuto 10 (OR7,071 IC95%1,269–39,415;p=0,014), infeccion materna intraparto (OR2,944 IC95%2,023–4,286;p=0,05) y choque neonatal (OR5,063 IC95%1,536–16,687;p=0,006). Habia aumento de riesgo con la presencia de mas de 6 antecedentes del parto o la adaptacion (OR3,188 IC95%2,124–4,78;p=0,006) o mas de 5 antecedentes postnatales (OR3,56 IC95%1,123–11,285;p=0,028). La alteracion hidroelectrolitica fue el unico hallazgo paraclinico identificado como factor de riesgo (OR 3,04 IC95% 0,974 – 11,934; p=0,05). Conclusion: Existen antecedentes perinatales, caracteristicas clinicas de las crisis y hallazgos paraclinicos que se comportan como factores de riesgo de mala respuesta a primera linea de tratamiento en crisis neonatales. La identificacion temprana de los pacientes en riesgo puede anticipar la necesidad de tratamiento adicional para prevenir las complicaciones asociadas a esta entidad. ABSTRACT RISK FACTORS ASSOCIATED WITH A POOR RESPONSE TO FIRST-LINE TREATMENT IN NEONATAL SEIZURES A CASE-CONTROL STUDY DONE AT MERCY FOUNDATION HOSPITAL Objective: To describe risk factors associated with poor response to phenobarbital as first-line anticonvulsant treatment in neonatal seizures at Mercy Hospital Foundation (Bogota, Colombia). Methods: Case-Control Study. We included patients younger than 44 weeks of conceptional age with neonatal seizures treated with phenobarbital. Controls (n=35) were patients with adequate response to phenobarbital. Cases (n=20) were those without response, which required a second-line anticonvulsant. Results: Seizure-related risk factors were: More than one seizure semiology (OR11,625 CI95%2,946–45,877;p<0,001), latency to phenobarbital startup greater than 12 hours (OR9,917 CI95%2,205–44,607;p=0,001), subtle (OR5,167 CI95%1,309–20,39;p=0,014) or tonic seizures (OR3,271 CI95%1,042–10,278;p=0,039),seizures longer than 5 minutes (OR2,824 CI95%1,927– 4,137;p=0,026). In perinatal history, the antecedents associated with an increase in risk were: low 10th minute Apgar score (OR7,071 IC95%1,269–39,415;p=0,014), maternal intrapartum infection (OR2,944 IC95%2,023–4,286;p=0,05), neonatal shock (OR5,063 CI95%1,536–16,687;p=0,006). There was a higher risk with more than 6 antecedents at birth/adaptation (OR3,188 IC95%2,124–4,78;p=0,006) or more than 5 postnatal antecedents (OR3,56 IC95%1,123–11,285;p=0,028). Abnormal electrolytes was the only finding at diagnostic tests identified as risk factor (OR 3,04 CI95% 0,974 – 11,934; p=0,05). Conclusion: There are perinatal antecedents, seizure features and diagnostic tests findings that behave as risk factors for therapeutic failure with phenobarbital as first-line anticonvulsant treatment in neonatal seizures. Early identification of patients at risk of poor response can anticipate the need for additional treatment to prevent complications associated with recurrent seizures.