This article addresses an integrative psycholinguistic and cognitive-semiotic perspective on constructing fear in English horror films. At the heart of constructing fear in horror film is the filmmakers’ presumption that viewers can potentially share their joint intention with the filmmakers, can share joint attention, and, as a result, share joint emotion. Drawing on the theory of intersubjectivity, fear in horror films emerges as the result of joint attention between filmmakers and viewers. Fear is viewed as a multimodal construct resulting from the synergistic integration of verbal, nonverbal, and cinematic semiotic resources via audial and visual modes. Each semiotic resource contributes to meaning-making by employing elements specific to horror films. The verbal system contains interjections, descriptive words, emotion-laden words, pleas for help, and violation of the sentence structure. The nonverbal elements include a contorted face, screaming, chaotic gestures, shaking, or stupor. The cinematic resource possesses the meaning-making potential to highlight various aspects of filmic fear through close-ups and middle close-ups, camera angles, dim light, and non-linear disturbing music. The meanings constructed by semiotic elements interact through cross-mapping, contributing to the formation of multimodal blends, which emerge in conceptual integration. Multimodal blends of fear in horror films include two-/three-component, non-parity, and consecutive patterns. From the viewers’ perspective, fear in horror films is perceived as a whole entity with a different level of intensity: from anxiety to horror. The experiment results show that the main indicators of fear for both males and females are pleas for help, voice and facial expressions, and music and close-up. However, while watching horror films, males feel interested more, while females experience negative emotions of fear, disgust, and tension.