Amyotrophic lateral sclerosis (ALS) research is advancing quickly, but the transition from phase II to phase III trials remains particularly challenging. In part, this is because of the paradox of phase II ALS trials - they are expected to inform researchers about safety, tolerability, dosage selection, and efficacy using a small number of patients, and relying on essentially the same outcome measures used in phase III trials. We examined pharmacokinetics in the cerebrospinal fluid and pharmacodynamic markers to demonstrate target engagement. In addition, primary outcome measures are often not specified or do not reach pre-specified goals for significance. We conclude that future trials should include pharmacokinetic (preferably in CSF) and pharmacodynamic markers of target engagement when possible. Primary endpoints should be pre-specified. Inclusion criteria should be used to reduce heterogeneity and target a relevant subpopulation of people with ALS when possible. Multiple phase II trials might be required before moving to a large phase III trial.