While the efficacy of the intracardiac defibrillators (ICDs) for primary prevention is not disputed, the relevant studies were carried out a long time ago. Most pertinent trials, including MADIT-II, SCD-Heft, and DEFINITE, recruited patients more than 20 years ago. Since then, improved therapeutic modalities including, in addition to cardiac resynchronization therapy, mineralocorticoid receptor antagonists, angiotensin receptor-neprilysin inhibitors, and, most recently, inhibitors of sodium-glucose cotransporter 2, have lowered present-day rates of mortality and of sudden cardiac death. Thus, nowadays, ICD therapy may be less effective than previously reported, and not as beneficial as many people currently believe. However, criteria for ICD implantation remain very inclusive. The patient must (only) be symptomatic and have ejection fraction (EF) ≤ 35%. The choice of EF 35% is notable because the average EF in all large trials was much lower, and clinical benefit was mainly limited to EF ≤ 30%. This EF cut-off value defines a substantial portion of potential ICD recipients. It seems therefore reasonable to limit ICD eligibility criteria in the EF range 30–35% to patients at highest risk only. We discuss and present some rational criteria to assist the clinician in improving risk stratification for preventive ICD implantation.