Deficits in facial emotional intensity recognition have been associated with social cognition in patients with major depression. The study examined multiple event-related potential (ERP) components in patients with major depression and investigated the relationships between ERPs, social cognition, and clinical features. Thirty-one patients met DSM-IV diagnosis of depression and 31 healthy participants completed the emotion intensity recognition task (EIRT), while ERPs were recorded. Data on ERP components (P100, N170, P200, and P300) were analyzed. The behavioral results showed that patients with major depression performed worse on EIRT, including all six categories of emotions (sadness, disgust, happiness, surprise, anger, and fear), compared to healthy participants. The ERP results showed that patients with major depression exhibited higher P100 amplitudes for sad and happy faces than healthy participants; P300 amplitudes induced by sad and surprise faces were also higher than in healthy participants, mainly in the central and temporal lobes. A positive correlation was found between sadness intensity scores and P100 amplitudes in patients with major depression. Patients with major depression are biased in their identification of facial expressions indicating emotional intensity. Specifically, they have emotional biases in the early and late stages of cognitive processing, mainly in the form of sensitivity to sad stimuli. It may lead to a persistent rumination of sadness that is detrimental to the remission of depression. Additionally, patients with major depression devote different amounts of cognitive resources for different intensities of sad faces during the preconscious stage of cognitive processing. The more intense their perception of sadness, the more cognitive resources they devote. Therefore, the assessment of the intensity of facial expressions is an important research topic, with clinical implications on social cognitive function in patients with major depression.