Commissioning is arguably the single most cost-effective strategy for reducing energy, costs, and greenhouse gas emissions in buildings today. Although commissioning has earned increased recognition in recent years, it remains an enigmatic practice whose visibility severely lags its potential. The application of commissioning to new buildings ensures that they deliver or exceed the performance and energy savings promised by their design and intended operation. When applied to existing buildings, commissioning identifies deficiencies and the almost inevitable "drift" from intended performance over time, and carries out interventions to put the building back on course. More formally, commissioning is a systematic, forensic approach to quality assurance and performance risk management, rather than a technology per se. This article presents the world's largest compilation and meta-analysis of commissioning experience and the associated literature, comprising 643 non-residential buildings, 99 million ft 2 of floorspace, $43 million in commissioning expenditures, and the work of 37 commissioning providers. The median normalized cost to deliver commissioning is $0.30/ft 2 ($2009 currencies) for existing buildings and $1.16/ft 2 for new construction (or 0.4% of the overall construction cost). The one third of projects for which data are available reveal over 10,000 energy-related deficiencies, the correction of which resulted in 16% median wholebuilding energy savings in existing buildings and 13% in new construction, with payback times of 1.1 and 4.2 years, respectively. Because energy savings exceed commissioning costs, the associated reductions in greenhouse gas emissions come at a "negative" cost of −$110/tonne CO 2 for new buildings and −$25/tonne for new construction. Cases with comprehensive commissioning attained nearly twice the overall median level of savings and five times the savings of the least-thorough projects. Significant non-energy benefits such as improved indoor air quality are also achieved. Applying the median whole-building energysaving values to the US non-residential buildings stock corresponds to an annual energy-saving potential of $30 billion (and 340 Mt of CO 2 ) by the year 2030.