Abstract Background The aims of this study were to identify the main variables associated with intra‐hospital mortality among patients transferred to a specialized neonatal care center, and to evaluate agreement and accuracy of referring and admission diagnoses. Methods A 6 month observational study was conducted to obtain clinical variables for intra‐hospital mortality among patients requiring interfacility transport. Association among variables was estimated using P oisson regression with robust variance. Agreement was evaluated between diagnosis before and after transfer using C ohen's kappa, sensitivity, specificity and Youden's I . Results The study included 191 neonates, 12.57% of whom died. Increased mortality was associated with the Transport Risk Index of Physiologic Stability ( TRIPS ). The associated variables were as follows: TRIPS (adjusted prevalence ratio [a PR] , 1.05; 95% confidence interval [ CI]: 1.02–1.08), weight 1500–2499 g (a PR , 0.08; 95% CI : 0.01–0.40), weight >2500 g (a PR , 0.56; 95% CI : 0.02–0.19), cardiopathy (a PR , 0.20 95% CI: 0.05–0.75), congenital defects (a PR , 4.59; 95% CI : 0.97–21.82) and renal failure (aPR, 3.69; 95% CI: 1.26–10.78). Diagnosis remained unchanged for 71.15% of the neonates. The greatest differences were hyaline membrane disease (49.4%) followed by transient tachypnea (59.9%). Youden's I for referral diagnosis was 0.22 for transient tachypnea, 0.66–0.69 for cardiopathy, esophageal atresia and pneumonia, 0.72–0.74 for hyaline membrane disease and pulmonary hypertension, and >0.90 for the remaining diagnoses. Conclusions Weight <1500 g, renal failure, congenital defects (except congenital cardiopathies) and high TRIPS were associated with a higher risk of intra‐hospital mortality. The findings suggest that improving transfer time and quality of care in ambulances would decrease mortality.