The purpose of the study was to assess frequency and predictors of disability measures in ALS. One hundred and fourteen newly diagnosed patients resident in eight administrative districts of Lombardy, Italy (population 4,947,554), included in a population-based registry, were followed for 2570 person-months (mean 22.5 months). The cumulative time-dependent risk of wheelchair, percutaneous endoscopic gastrostomy, and assisted ventilation was estimated according to the Kaplan-Meier method. Predictors of disability (age, sex, disease duration at diagnosis, type of onset, El-Escorial diagnosis) were assessed with the Cox proportional hazard function. During follow-up, 29 patients (25.4%) became wheelchair bound, 51 (44.7%) received gastrostomy, and 47 (41.2%) received assisted ventilation. The median time to loss of ambulation was 46.7 months (95% CI 36.5-56.8). The median time to gastrostomy and assisted ventilation was 31.1 months (95% CI 26.8-35.4) and 34.6 months (95% CI 29.6-39.6), respectively. Spinal onset ALS was the only predictor of loss of ambulation. Predictors of gastrostomy were older age, definite ALS, and shorter disease duration. Shorter disease duration was the only predictor of assisted ventilation. In conclusion, patients with ALS differ in terms of measures and predictors of disability. These factors are sources of bias and confounding in randomized clinical trials.