PurposeTo compare percutaneous radiologic gastrostomy (PRG) and percutaneous endoscopic gastrostomy (PEG) in terms of tolerance, efficacy, and survival in patients with amyotrophic lateral sclerosis (ALS). Materials and Methods Forty patients with ALS (17 men; mean age, 66.1 years; range, 39–83 y) underwent 21 PEG and 22 PRG attempts (including three unsuccessful PEG attempts) from 1999 to 2005. To assess tolerance and efficacy, a successful and well tolerated placement was defined as any successful placement with no major or minor local complications or pain requiring opioid analgesic agents. Univariate analysis was performed for all recorded parameters, followed by multivariate analysis for successful and well tolerated placement, 6-month mortality rate, and survival. Results General success rates were 85.7% for PEG and 100% for PRG. Pain was more frequent in PRGs (81.8% vs 52.4%; P = .05). Successful and well tolerated placement was seen in 81.8% of PRGs and 57.1% of PEGs (P = 0.1). Advanced age (P = .02) and PRG (P = .07) were predictive of successful and well tolerated placement. The interval from diagnosis to placement (P = .001) and ability to perform spirometry (P = .002) were predictive of survival. Oximetry measurements (P = .007) and interval from diagnosis to placement (P = .02) were predictive of mortality at 6 months. Conclusions PRG is more efficacious and better tolerated than PEG, essentially because it avoids the respiratory decompensation that may occur in PEG. Therefore, PRG should be preferred in cases of ALS. Survival is linked to ALS evolution and not to the choice of PRG or PEG placement.