The objective of this study was to identify prognostic factors for survival in amyotrophic lateral sclerosis from a large prospective observational study performed in France. The study included a cohort of 2069 patients fulfilling broad entry criteria treated with riluzole. Over 100 demographic, biological, clinical and quality-of-life variables were monitored and assessed for their effect on survival. Patients were randomized post hoc into two groups: one group (two-thirds of the patients) to generate the prognostic models and one group (one-third of the patients) to validate the resulting models. Thirteen variables were found to affect survival independently and were used to construct a survival prediction score, RL401. These included age, disease duration, slow vital capacity, intensity of tiredness (visual analogue scale), number of body levels with spasticity, atrophy and/or fasciculations, cough, distal muscle strength, household income, depression and two biological parameters, plasma creatinine levels and neutrophil counts. A simplified score, RL401S, was constructed, designed to be easy to use and interpret. The predictive powers of the two scores were similar.