Background: Patients with left ventricular systolic dysfunction (LVSD) are at high risk of sudden cardiac death (SCD). Implantable cardioverter defibrillators (ICDs) have an important role in preventing SCD in selected patients with LVSD and chronic heart failure (CHF). Drug therapies for LVSD and CHF also appear to also be useful in reducing SCD. However, the magnitude of benefit of these approaches on SCD is uncertain. We therefore conducted a meta-analysis comparing the effect on SCD achieved by ICDs versus medical therapies, additional to standard background medical therapies including ACE inhibitors and/or beta-blockers (BBs). Methods: Our meta-analysis included trials of > 100 patients with reduced left ventricular ejection fraction (LVEF), i.e., <40 . Fourteen randomized controlled trials met the criteria for meta-analysis, 10 involving medical therapies (angiotensin receptor blockers [ARBs], mineralocorticoid receptor antagonists [MRAs], ivabradine, n3-polyunsaturated fatty acid [PUFA], ferric carboxymaltose and aliskiren) and four involving ICDs. Results were pooled using the Mantel-Haenszel random effects method. Results: Drug therapy (n = 36,172) reduced the risk of SCD overall (risk ratio (RR) = 0.89, 95 confidence interval (CI) = 0.82-0.98, p = 0.02) when compared to placebo. MRAs alone were most effective in reducing SCD (n = 11,032, RR = 0.79 [0.68-0.91], p = 0.001). ICD insertion greatly reduced SCD (n = 4,269, RR = 0.39 [0.30-0.51], p <0.00001) compared with placebo. The difference in treatment effect between the ICD and drug therapy was significant (p <0.002), and between ICD and MRAs (p <0.002). Conclusions: Drug therapies when added to a standard background regimen comprising ACE inhibitor and/or BB reduced SCD overall and MRAs alone were most effective in this regard. ICDs were more effective than drugs in SCD abrogation. However, the added procedural morbidity and the cost of ICD need to be considered in decision-making re-approach to SCD reduction in the individual patient.