The study evaluated whether general practitioners, after brief training, can use point-of-care ultrasound (PoCUS) to improve the diagnosis of heart failure in patients with new onset of dyspnea in primary care. They enrolled 112 adult patients with unexplained dyspnea, of whom 102 underwent standardized hand-held lung and cardiac ultrasound where B-lines were quantified and left ventricular ejection fraction (LVEF) was assessed. The diagnoses before and after PoCUS were compared with the cardiologist's final diagnosis based on standard echocardiography. After PoCUS, sensitivity increased to 86.8%, specificity to 88.2%, positive predictive value to 93.7%, and negative predictive value to 76.5%, compared to clinical judgment before PoCUS (sensitivity 5.3%, specificity 73.4%, negative predictive value 56.5%). Agreement with final diagnosis improved substantially (Cohen's κ from 0.254 to 0.723). Physician-rated B-lines and LVEF were strongly correlated with cardiologist results. Short training enables GPs to effectively use PoCUS to detect heart failure, improving diagnostic accuracy and reducing unnecessary referrals.