<p>Despite the proliferation of asynchronous video instruction across higher education and K–12 contexts, educational technology lacks an integrated, design-oriented framework that explicitly addresses how cinematic composition choices influence the mechanisms through which learning occurs. This conceptual paper introduces Cinematic Instructional Design (CID), a provisional integrative framework that synthesizes insights from multimedia learning, instructor-presence research, multimodal communication, and film and media theory. Rather than treating cinematic form as pedagogy in itself, CID conceptualizes framing, sound design, editing rhythm, narrative structure, and visual signaling as design variables that may shape attention allocation, perceived social presence, affective engagement, and cognitive load, mechanisms that, in turn, may support or impede learning in asynchronous environments. Drawing on four decades of empirical research, CID articulates eight interrelated design dimensions, each linked to hypothesized causal mechanisms and operationalizable indicators. Unlike procedural production rubrics or purely cognitive-load frameworks, CID attends to the visual-compositional and social-affective dimensions of instructional video that existing frameworks address incompletely. The framework is offered as a provisional, design-oriented synthesis rather than a validated theory: its value lies in clarifying a neglected design problem, specifying plausible mechanisms, and generating testable propositions for future design-based research, with particular attention to equity-relevant outcomes. The framework synthesizes research-based design principles drawn from multimedia learning, instructor-presence, multimodal communication, and narrative cognition research into a structured vocabulary that practitioners can apply and researchers can test.</p>