OBJECTIVES: The aims were to analyse changes in nutritional parameters from diagnosis of amyotrophic lateral sclerosis (ALS) to death and to assess their relationships with survival at the time of diagnosis and during follow-up. METHODS: 92 ALS patients were included and clinically assessed every 3 months (ALS functional rating scale, manual muscular testing, forced vital capacity, weight, BMI, percentage weight loss). Bioimpedance was performed to evaluate body composition (fat-free mass, fat mass and hydration status) and phase angle. Survival analyses were performed from diagnosis to death or censoring date using a Cox model. RESULTS: The evolution of nutritional parameters in ALS patients was marked by significant decreases in weight, BMI, fat-free mass and phase angle, and increased fat mass. The authors identified an adjusted 30% increased risk of death for a 5% decrease from usual weight at time of diagnosis (RR 1.30; 95% CI 1.08 to 1.56). During follow-up, the authors identified adjusted 34% (95% CI 18% to 51%) and 24% (95% CI 13% to 36%) increased risks of death associated with each 5% decrease in usual weight and each unit decrease in usual BMI, respectively (p<0.0001). Malnutrition during the course was related to a shorter survival (p=0.01), and fat mass level was associated with a better outcome (RR 0.90 for each 2.5 kg fat mass increment). CONCLUSIONS: Nutritional parameters of ALS patients worsened during evolution of the disease, and worse nutritional status (at time of diagnosis or during the course) was associated with a higher mortality. This study offers some justification for studying the use of therapeutic nutritional intervention to modify the survival of ALS patients.