Most studies on sensory extinction have focused on selected patients with subacute and chronic right hemisphere lesions. In studies conducted on acute stroke patients, risk factors and time course were not evaluated. Our aim was to determine the prevalence, risk factors, and time course of sensory extinction in the acute stroke setting. Consecutive patients with acute stroke were tested for tactile, visual, auditory, and auditory-tactile cross-modal extinction, as well as for peripersonal visuospatial neglect (PVN). Tests were repeated at 2, 7, 15, 30, and 90 days after initial examination. A multivariable logistic regression analysis was performed to test the association between sensory extinction and demographic and clinical risk factors. Seventy-three patients (38.4% women) were recruited: 64 with ischemic stroke and nine with haemorrhagic stroke. Mean age was 62.3 years (95% CI 58.8-65.7), mean NIHSS score was 1.6 (95% CI 1.2-2.1), and mean time to first examination was 4.1 days (95% CI 3.5-4.8). The overall prevalence of all subtypes of sensory extinction was 13.7% (95% CI 6.8-23.8). Tactile extinction was the most frequent subtype with a prevalence of 8.2% (95% CI 3.1-17.0). No extinction was found beyond 15 days after the first examination. After adjustment for age, sex, lesion side, type of stroke, time to first examination and stroke severity, a lesion volume ≥2 mL (adjusted OR = 38.88, p = 0.04), and presence of PVN (adjusted OR = 24.27, p = 0.04) were independent predictors of sensory extinction. The insula, the putamen, and the pallidum were the brain regions most frequently involved in patients with sensory extinction. Extinction is a rare and transient phenomenon in patients with minor stroke. The presence of PVN and lesion volume ≥2 mL are independent predictors of sensory extinction in acute stroke.