Pupillary Examination & the Swinging Flashlight Test (RAPD)

The pupillary exam, and specifically the swinging flashlight test, detects a relative afferent pupillary defect (RAPD) — an asymmetry in the afferent (sensory) visual pathway between the two eyes. Because both pupils normally constrict equally regardless of which eye is stimulated (the consensual reflex), the swinging flashlight test isolates afferent pathway dysfunction (optic nerve or severe, extensive retinal disease) from efferent (motor, third-nerve or iris) abnormalities, which affect only the ipsilateral pupil's direct response.

The room is dimmed, and the patient fixates on a distant target to minimize a near-response confound. The examiner shines a bright light on one pupil for three to five seconds, observing the direct response, then swiftly swings the light to the other eye, moving quickly across the bridge of the nose to minimize the time both pupils spend in darkness. This back-and-forth cycle is repeated several times, comparing each pupil's response to direct illumination. A normal result shows equal, brisk constriction of each pupil regardless of which eye is illuminated. An abnormal, positive RAPD occurs when the light swings to the affected eye and that pupil appears to paradoxically dilate, or constrict less briskly, compared to the other eye, since its afferent signal is weaker than the consensual signal it had just been receiving. When present, an RAPD can be graded using neutral density filters of increasing strength placed in front of the unaffected eye until the apparent defect is neutralized, giving a semi-quantitative severity estimate.

A positive RAPD (Marcus Gunn pupil) classically results from unilateral or markedly asymmetric optic neuropathy — optic neuritis, ischemic optic neuropathy, or a compressive lesion — or from extensive unilateral retinal disease such as a large retinal detachment or central retinal artery or vein occlusion. Its absence despite unilateral vision loss suggests the loss is not due to afferent pathway dysfunction on that side, but rather to a refractive, media, or bilaterally symmetric process, or a non-organic cause — an important localizing clue in its own right.

Common mistakes include swinging the light too slowly, which allows both pupils time to re-dilate in darkness between stimulations and obscures the direct comparison the test relies on; testing in a room that is too bright or too dark; failing to have the patient fixate on a distant target; and missing a subtle RAPD by not swinging the light enough times or not watching closely enough for a partial rather than fully paradoxical response.

A positive RAPD is one of the most objective, reliable bedside signs in neuro-ophthalmology, since it does not depend on subjective patient report and cannot be feigned or reliably suppressed. It reflects afferent pathway asymmetry rather than visual acuity per se — a patient can have severely reduced acuity in one eye from a refractive or media cause without any RAPD, while a relatively mild but asymmetric optic neuropathy can produce a clearly positive RAPD. Emerging cell phone-based and automated pupillometry tools show promise for RAPD screening by non-specialists, with one validated application achieving high sensitivity (93%) and specificity (85%) against neuro-ophthalmologist assessment.¹ However, objective pupillometry without careful methodological control can show substantial measurement variability even in healthy individuals, generating potential false positives, reinforcing that automated tools currently supplement rather than replace skilled manual testing.²

A detected RAPD should always be correlated with the rest of the neuro-ophthalmic exam — visual acuity, color vision, visual fields, and fundus exam — to localize the underlying cause, and an automated pupillometry score should not be treated as diagnostic on its own without correlation to careful manual testing.