Introduction: The evaluation of health programs effectively contribute to the planning and direction of public health policies. This study aimed to evaluate a surveillance program to the newborn at risk. Objective: Cross-sectional analytical study with data collected in 2009, from Maringá-PR. Method: We extracted maternal, infant and care information from a stratified sample of records and monitoring reports of children born in 2008, totaling 250 newborns at risk. Data were analyzed by chi-square test and multiple logistic regression. Results: Of the newborns at risk included in the program, 37% had satisfactory monitoring. It was found that assistance was a spontaneous demand with an inefficient access of risk priority. Low birth weight (OR=2.30; IC95%=1.25-4.23), coupled with the insufficient number of visits (OR=7.11; IC95%=2.34-21.63), guidelines (OR=2.49; IC95%=1.24-5.01) and weighing (OR=2.05; IC95%=1.01-4.15), contributed to the inappropriate monitoring of children at risk by the program. Conclusion: From the results found on the attention to the child at risk, it is suggested recast proposals of the inclusion criteria and strategies that prioritize a scheduled and more consonant assistance with the recommendations of the Ministry of Health.