The requirement for those providing healthcare to measure outcomes raises the fundamental question of how these should be measured. Both the scale of the low vision population and the different models of care proposed make it a pressing need to determine outcomes using valid and reliable measures. In this study we describe the development and piloting of a questionnaire designed to meet this requirement. The questionnaire was piloted on 56 subjects with age-related macular degeneration (ARMD). Test-retest reliability of the questionnaire was estimated using a further sample of 28 subjects with ARMD. Pilot testing confirmed the feasibility of administering the questionnaire to an elderly population with ARMD. A total of 87% of the sample reported using an LVA (65% on a daily basis), predominantly for a reading task. The number of tasks an LVA was used for, its frequency of use, length of continuous use and ease of use are significantly associated with the rating of an LVA as important/less important (P < 0.0001), but the acuity achieved with the device is not significantly associated with rated importance (P = 0.11), reinforcing doubt about the validity of visual function outcomes. The limitations of this pilot study and earlier case series in describing the outcomes of low vision care are discussed and a randomised controlled trial using a broad range of outcomes is proposed in order to fully describe the effectiveness of low vision rehabilitation and inform the debate about models of care.