Minor potential cardioembolic sources of stroke such as atrial septal aneurysms (ASA) or patent foramen ovale (PFO) are important risk factors for cryptogenic stroke. We aim to determine the prevalence of these abnormalities through an exhaustive etiological workup including transesophageal echocardiography and cervical arteries assessment in stroke patients younger than 60 years of age who had no evidence of a significant source of embolism. We classified 118 stroke patients into four groups according to transesophageal echocardiography (TEE) and cervical arteries assessment findings. Group A, consisted of 30 (25.4%) patients who had an arteriopathy likely related to stroke without any cardiac abnormality; Group B, 49 (41%) patients who had only a potential cardiac source; Group C, 9 (7.6%) patients who had an obvious arterial source of stroke and incidental cardiac abnormalities, and Group D, 30 (25.4%) patients who had neither cardiac nor arterial source. Data were analysed with X2 test for the comparison of risk factors between groups. Variance analysis was used to compare age between groups. Significance was assessed as p < 0.05. ASA represented 56.8% of the cardiac abnormalities and was diagnosed in 35.4% of the 79 patients who had an unexplained stroke (B and D). A PFO was found in 34.1% of the patients who had a cryptogenic stroke (B and D). According to Fisher's exact test, ASA was significantly associated to PFO (p << 0.001). According to this selection one fourth of the patients might have a truly cryptogenic stroke as the etiological workup failed to demonstrate any source of stroke. Comparison between groups showed that the patients in whom an arterial source was detected also had a potential cardioembolic source in 23% of the cases (C), versus 62% in patients who had no arterial source (B and D) (p = 0.0007). Our study confirmed the strong association between ASA, PFO and stroke. Although there was a lower incidence of minor potential cardioembolic sources in patients who had a cervical artery disease, we suggest a systematic TEE screening in all patients with stroke without major cardiac source, in order to ensure a better prevention.