Background: Differentiating between benign and malignant pulmonary nodules is a diagnostic challenge, and inaccurate detection can result in unnecessary invasive procedures. Cell-free DNA (cfDNA) has been successfully utilized to detect various solid tumors. In this study, we developed a genome-wide approach to explore the characteristics of cfDNA sequencing reads obtained by low-depth whole-genome sequencing (LD-WGS) to diagnose pulmonary nodules. Methods: LD-WGS was performed on cfDNA extracted from 420 plasma samples from individuals with pulmonary nodules that were no more than 30 mm in diameter, as determined by computed tomography (CT). The sequencing read distribution patterns of cfDNA were analyzed and used to establish a model for distinguishing benign from malignant pulmonary nodules. Results: We proposed the concept of weighted reads distribution difference (WRDD) based on the copy number alterations (CNAs) of cfDNA to construct a benign and malignant diagnostic (BEMAD) algorithm model. In a training cohort of 360 plasma samples, the model achieved an average area under the receiver operating characteristic (ROC) curve (AUC) value of 0.84 in 10-fold cross-validation. The model was validated in an independent cohort of 60 plasma samples, obtaining an AUC value of 0.87. The BEMAD model could distinguish benign from malignant nodules at a sensitivity of 74% and a specificity of 86%. Furthermore, analysis of the critical features of the cfDNA using the BEMAD model identified repeat regions that were associated with microsatellite instability, which is an important indicator of tumorigenesis. Conclusions: This study provides a novel non-invasive diagnostic approach to discriminate between benign and malignant pulmonary nodules to avoid unnecessary invasive procedures.