We describe designs for clinical trials in ALS including two that are more efficient than the standard two-arm, parallel design. The more efficient designs incorporate a lead-in period followed by a randomized intervention (drug or placebo) period. Efficacy of the more efficient designs is based on measuring, within each patient, the difference in slope while on the new treatment compared to the lead-in period. We demonstrate, with sample size calculations, that the lead-in designs are considerably more efficient than the standard two-arm, parallel design. Sample sizes can be reduced by 44% for a 12-month study using ALSFRS rate of decline as a primary endpoint for a two-arm trial with 4 months lead-in compared to a parallel design. A sample size reduction of 70% can be realized with variable lead-in compared to a parallel design.