Background: AMX0035 is an oral fixed-dose coformulation (sodium phenylbutyrate-taurursodiol) that significantly slowed functional decline as measured by the Amyotrophic Lateral Sclerosis Functional Rating Scale–Revised (ALSFRS-R) in a 24-week randomized, placebo-controlled trial in ALS (CENTAUR). Participants completing CENTAUR were eligible to enroll in an open-label extension (OLE) and receive AMX0035 (≤ 132 weeks). In a nearly 3-year analysis of all randomized participants from CENTAUR, those originally randomized to AMX0035 had a 6.5-month longer median survival compared to those originally randomized to placebo. Objective: To report the risk of key study events (death, permanent assisted ventilation [PAV; defined as >22 hours/day for >7 days], and first hospitalization) through July 2020 (longest post-randomization follow-up, 35 months) among all participants randomized in CENTAUR (N=137). Time to death was captured by OmniTrace via a search of public records for all participants including those who discontinued, were lost to follow-up, or did not enroll in the OLE. Other key events were recorded prospectively via clinic reports, with censoring at last date of follow-up for those without reported events. Hazard ratios (HRs) were estimated using a Cox proportional hazards model with covariates of prebaseline ALSFRS-R slope, baseline ALSFRS-R, and age. Results: The mean hazard of any key event was 44% lower in those originally randomized to AMX0035 versus placebo (HR, 0.56; 95% CI, 0.37–0.86; P=0.008); median times to event were 16.3 (95% CI, 13.5–20.3) and 13.8 (95% CI, 11.5–16.6) months, respectively. The mean hazard of first hospitalization was 44% lower in the group originally randomized to AMX0035 versus placebo (HR, 0.56; 95% CI, 0.32–0.96; P=0.034); median times to event were not reached (NR) (95% CI, 14.8–NR) and 14.4 (95% CI, 6.8–NR) months, respectively. The mean hazard of death or PAV was 42% lower in those originally randomized to AMX0035 versus placebo (HR, 0.58; 95% CI, 0.36–0.93; P=0.025), with median times to event of 23.8 (95% CI, 18.2–29.1) and 18.5 (95% CI, 13.5–21.7) months, respectively. Conclusions: Long-term risk of death or PAV and risk of first hospitalization was significantly lower among those originally randomized to AMX0035 versus placebo. Though limited by the potential for missing data due to loss to follow-up, these results supplement the previously reported functional and long-term survival results seen with AMX0035.