While considerable emphasis has been put on investigating the mechanisms that drive reduced social connection in patients with schizophrenia (SCZ), recent studies have increasingly focused on the issue of loneliness in SCZ. Both social cognitive bias and self-reported empathy predict loneliness in non-clinical populations, the current study, therefore, aims to examine the relationship between loneliness, reduced social connection and social cognitive biases, and self-reported empathy in SCZ. Ninety-three adult SCZ and sixty-six matched healthy individuals completed a battery of questionnaires measuring loneliness and social connection (Revised-UCLA Loneliness scale, Lubben Social Network Scale, Social Disconnectedness Scale), cognitive biases (Ambiguous Intentions Hostility Questionnaire (AIHQ); Davos Assessment of Cognitive Biases Scale (DACOBS), Cognitive Biases Questionnaire for psychosis, CBQp) and self-reported empathy (Interpersonal Reactivity Index, IRI). Significant predictors of loneliness in SCZ were entered into two latent variables (“Social Threat Bias”, “Social Connection”), and structural equation modeling was used to explore the direct and indirect relationships between Social Threat Bias and loneliness in SCZ. Patients reported higher levels of loneliness, cognitive biases, and personal distress compared to controls. Furthermore, SCZ reported less social connection and perspective-taking compared to controls. AIHQ Blame Score, CBQp Threatening Events, DACOBS, and IRI Personal Distress were significantly associated with loneliness in SCZ. SEM modeling revealed that Social Threat Bias was linked to increased loneliness in SCZ both directly and indirectly via a decreased social connection. The results of the current study suggest that social threat bias should be considered while planning the interventions aimed to reduce loneliness in schizophrenia.