Earlier studies provided preliminary support for the role of classical conditioning as a pathway of disgust learning, yet this evidence has been limited to self-report. This study included facial electromyographical (EMG) measurements (corrugator and levator muscles) and a behavioural approach task to assess participants' motivation-to-eat the actual food items (conditioned stimuli, CS). Food items served as CS and film excerpts of a woman vomiting served as unconditioned stimuli (US). Following acquisition the CS+ (neutral CS paired with US disgust) was rated as more disgusting and less positive. Notably, the conditioned response was transferred to the actual food items as evidenced by participants' reported lowered willingness-to-eat. Participants also showed heightened EMG activity in response to the CS+ which seemed driven by the corrugator indexing a global negative affect. These findings suggest that classical conditioning as a pathway of disgust learning can be reliably observed in subjective but not in disgust-specific physiological responding.