The Patient-Ranked Order of Function (PROOF) is a novel approach to account for patient-reported preferences in the evaluation of treatments of amyotrophic lateral sclerosis (ALS). In this study, we assess the reliability and prognostic value of different sets of patient-reported preferences that can be used for the PROOF end point. Data were obtained through online surveys over the course of 12 months using the population-based registry of the Netherlands. Patients were asked to score functional domains of the ALS Functional Rating Scale (ALSFRS-R) and rank the order of importance of each domain. Two weeks after the initial invite, the questionnaire was repeated to evaluate test-retest reliability. Vital status was extracted from the municipal population register. In total, 611 patients with ALS were followed up for survival and 382 patients were included in the test-retest reliability study. All versions of PROOF, using different sets of preferences, resulted in excellent reliability (intraclass correlation coefficients ranged from 0.89 [95% CI 0.87-0.91] to 0.97 [95% CI 0.97-0.98], all p < 0.001), without systematic differences between baseline and week 2 (mean rank difference range -1 to -3 [95% CI range -8 to 2], all p > 0.20). Preferences about future events were more variable than preferences about current symptoms. All versions of PROOF strongly predicted overall survival (hazard ratios per 10th rank percentile ranged from 0.80 to 0.83 [95% CI range 0.76-0.87], all p < 0.001) and had a more even separation of survival curves between rank-stratified subgroups compared with the ALSFRS-R total score. In a large cohort of patients, we show how patient-reported preferences can be measured and integrated reliably with the ALSFRS-R without leading to systematic bias. Patient preferences may provide unique prognostic information in addition to what is already measured conventionally. This could provide a more comprehensive understanding of how medical interventions effectively address the patient's concerns and improve what matters most to them.