Background: Neonatal jaundice resulting from raised blood bilirubin levels is one of the most common clinical conditions that needs medical attention. To initiate appropriate management that can both prevent and treat severe neonatal jaundice, screening methods that measure bilirubin level are warranted. Methods: In this study, we present an Android OS-based application for detecting neonatal jaundice. We used the application to detect jaundice in 113 neonates. Our smartphone-based estimation of bilirubin levels depends on a smartphone, a color calibration card, and a 100X zoom microscope clip. Our application was designed to acquire images of the newborn's forehead skin in a standardized manner, estimate the average R, G, B scores of the images that have been taken from the forehead skin and calibration card, and then convert them to hue, saturation, intensity (HSI) parameters. All these are performed offline; in this application, we used offline machine learning and regression techniques for analysis Results: Our smartphone-based estimation of bilirubin levels had a sensitivity of 68% and specificity of 92.3% for estimating the bilirubin levels of less than 10 mg/dL and sensitivity of 82.1% and specificity of 100% for estimating the bilirubin levels of less than 15 mg/dL. Our application-based estimation of bilirubin levels had the correlation of 0.479 with the total serum bilirubin values. Conclusions: Our results suggest that our smartphone-based application can serve as a promising screening tool for neonatal jaundice, and it can aid in determining neonates requiring a blood draw for measuring total serum bilirubin level.