HIV prevalence worldwide among people who inject drugs (PWID) is around 19%. Harm reduction for PWID includes needle-syringe programs (NSPs) and opioid substitution therapy (OST) but often coupled with antiretroviral therapy (ART) for people living with HIV. Numerous studies have examined the effectiveness of each harm reduction strategy. This commentary discusses the evidence of effectiveness of the packages of harm reduction services and their cost-effectiveness with respect to HIV-related outcomes as well as estimate resources required to meet global and regional coverage targets. NSPs have been shown to be safe and very effective in reducing HIV transmission in diverse settings; there are many historical and very recent examples in diverse settings where the absence of, or reduction in, NSPs have resulted in exploding HIV epidemics compared to controlled epidemics with NSP implementation. NSPs are relatively inexpensive to implement and highly cost-effective according to commonly used willingness-to-pay thresholds. There is strong evidence that substitution therapy is effective, reducing the risk of HIV acquisition by 54% on average among PWID. OST is relatively expensive to implement when only HIV outcomes are considered; other societal benefits substantially improve the cost-effectiveness ratios to be highly favourable. Many studies have shown that ART is cost-effective for keeping people alive but there is only weak supportive, but growing evidence, of the additional effectiveness and cost-effectiveness of ART as prevention among PWID. Packages of combined harm reduction approaches are highly likely to be more effective and cost-effective than partial approaches. The coverage of harm reduction programs remains extremely low across the world. The total annual costs of scaling up each of the harm reduction strategies from current coverage levels, by region, to meet WHO guideline coverage targets are high with ART greatest, followed by OST and then NSPs. But scale-up of all three approaches is essential. These interventions can be cost-effective by most thresholds in the short-term and cost-saving in the long-term.