Abstract Purpose To distinguish healthy infants from potential lung disease patients immediately after birth using lung ultrasound (LUS). Design, Setting and Patients This is a nested case-control study containing 22 lung disease patients and 473 healthy infants from a total of 504 consecutive infants. The infants were admitted to the Obstetrics & Gynecology Hospital of Fudan University, Shanghai, China, from 1 January 2020 to 1 April 2020. A newly designed scanning protocol was used to capture LUS images. The sensitivity, specificity, PPV and NPV for predicting healthy infants and patients were calculated individually. The transition process image patterns and their variations are shown. The relationship between clinical signs and high-risk image patterns was calculated by Kendall’s tau-b test. Measurements and main results LUS images were captured, and their predictive value was calculated. Four low-risk patterns could typically be seen only in healthy infants (specificity=86.4%, PPV=99.0%), whereas four high-risk patterns could be seen in both healthy infants and patients (specificity=62.4%, PPV=9.6%). High-risk patterns were more likely to be pathological signs when appearing at the oxter and lower back and physiological signs when appearing at the prothorax. These high-risk patterns are significantly related to clinical signs. All these patterns are consistent during the first 6 hours after birth. Conclusions LUS is a valid modality for differentiating healthy infants from potential patients with mild respiratory difficulty. Four low-risk patterns had high value in predicting healthy infants, but four high-risk patterns were not specific enough to discover lung disease patients.