PURPOSE: To develop a questionnaire for assessing self-reported visual problems under low luminance and at night for use in studies on age-related maculopathy (ARM). METHODS: The questionnaire was developed in three steps: (1) Content for questionnaire items was identified through focus groups of older adults with ARM and those exhibiting normal retinal aging. The topic for discussion was "vision at night and under low lighting." Discussion was audiotaped, transcribed, and subjected to content analysis to identify problem categories expressed by patients. (2) This content was used to develop a preliminary questionnaire administered by telephone to persons with ARM or normal retinal aging. Principal-components analysis identified groups of items that formed the questionnaire's subscales that were evaluated for internal consistency, and an item-reduction strategy was implemented to generate a briefer questionnaire. (3) Psychometric properties of the shortened Low-Luminance Questionnaire (LLQ) were determined, including construct validity, criterion validity, and test-retest reliability. RESULTS: The 32-item LLQ has six subscales (driving, extreme lighting, mobility, emotional distress, general dim lighting, and peripheral vision), all with good internal consistency (Cronbach alpha > or = 0.82). Scores on LLQ subscales correlated moderately with nearly all National Eye Institute Visual Function Questionnaire (NEI VFQ)-25 subscales and decreased in value (indicating more disability) for patients with increasing ARM disease severity. Whereas rod-mediated parameters of dark adaptation were significantly associated with LLQ subscale scores (r = 0.19-0.43, all P < 0.03), cone-mediated parameters were not. Test-retest reliability ranged from 0.74 to 0.88 for all subscales (P < 0.0001), except for peripheral vision (0.46; P = 0.0003), which also exhibited a ceiling effect in almost half of the respondents. CONCLUSIONS: The 32-item LLQ, derived from the content of focus group comments by persons with ARM, has good construct validity, subscale scores related to rod-mediated visual function, and good test-retest reliability for five of six subscales. The LLQ may ultimately be useful in patient-centered evaluation of the outcome of interventions to prevent ARM or to arrest progression of early disease.