Do carbon offsets actually work? Some do, but most on the voluntary market don’t, and the difference comes down to the specific project, the standard behind it, and how honestly the underlying emissions reduction was measured. Research has repeatedly found that offset programs overstate their climate impact, sometimes by a factor of ten or more, while a narrower slice of high-integrity credits — particularly methane capture and engineered carbon removal — can deliver measurable benefit. Whether the credit in front of you is one of those is a question you have to answer project by project.
What the Evidence Actually Shows
The most damaging assessment of the voluntary market came in January 2023, when The Guardian, Die Zeit, and the research group SourceMaterial analyzed peer-reviewed studies of rainforest projects certified by Verra, which operates the largest crediting program in the voluntary market. The investigation concluded that roughly 94 percent of the credits produced by those projects should not have been approved. Only 8 of the 29 projects where analysis was possible showed evidence of meaningful deforestation reductions. On average, the threat to the forests had been overstated by about 400 percent; excluding three unusually strong Madagascar projects, the average inflation rose to roughly 950 percent. Verra disputed the methodology and announced reforms.
Researchers at the University of Oxford’s Smith School reached a similar conclusion in blunter terms, finding that offset programs have “failed for 25 years” and that prior research shows they “routinely overestimate their climate impact, in many cases by as much as a factor of ten or more.”
Nature-based credits also face a durability problem that is getting worse, not better. California’s forest offset program requires developers to deposit roughly 20 percent of generated credits into a buffer pool that acts as an insurance reserve against fire, disease, and other losses.1CarbonPlan. Nearly a Third of California’s Forest Offset Buffer Pool Has Gone Up in Smoke Recent wildfire seasons have consumed a significant share of that reserve, raising the question of whether the current buffer levels come close to pricing wildfire risk in a warming climate.
The failures cluster in a few predictable places: avoided-deforestation projects where the baseline threat was inflated, renewable energy credits in markets where clean power was already the cheapest option, and any project relying on the counterfactual “this would have happened without us.” The gap between the best and worst credits is enormous, and understanding why credits fail is the fastest way to spot the ones worth buying.
The Four Tests Every Credit Has to Pass
A carbon credit represents one metric ton of carbon dioxide, or equivalent, reduced or removed from the atmosphere.2UNDP Climate Promise. What Are Carbon Markets and Why Are They Important Whether that ton is real depends on four criteria. When credits fail, one or more of them was compromised.
Additionality
A credit is legitimate only if the reduction would not have happened without the carbon revenue. If a forest was never truly going to be logged, paying to “protect” it produces credits with zero climate value. If a wind farm would have been built anyway on its own economics, the credit didn’t cause the reduction. Evaluating additionality means comparing the project against common practice in its region and asking whether the project pencils out without carbon income. This is where the biggest integrity failures happen, because the incentive to overstate the threat to a forest, or to understate a project’s baseline profitability, is enormous when every phantom ton can be sold for real money.
Permanence
Carbon stored by a project has to stay out of the atmosphere long enough to matter. Most standards define permanence as at least 100 years, aligned with the IPCC’s 100-year time horizon for global warming potential.3Climate Action Reserve. Keeping It 100 – Permanence in Carbon Offset Programs Nature-based projects carry the greatest reversal risk. Forests burn, get diseased, or are logged by a future landowner. Standards attempt to manage this through buffer pools, but as the California experience shows, those reserves may not be sized for the fire regimes now occurring.
Leakage
Leakage happens when reducing emissions in one place just pushes them somewhere else. Protecting one tract of forest is worthless if the logging company moves next door and cuts the same volume of timber. Developers are supposed to model and subtract this displacement. Research has found that leading certification standards significantly underestimate leakage, introducing what one analysis called a risk of “silent failure” in nature-based programs.4Grantham Research Institute on Climate Change and the Environment. Avoiding Leakage from Nature-Based Offsets by Design
Double Counting
Two parties cannot both claim the same reduced ton. Within a registry, serial numbers and retirement records prevent this at the project level.5Climate Action Reserve. Serial Number Guide Across countries it gets harder. Under Article 6 of the Paris Agreement, when one country sells credits to another, the seller is supposed to apply a “corresponding adjustment” so the reduction doesn’t count toward both nations’ climate pledges.6UNFCCC. Article 6 of the Paris Agreement Those rules are still being implemented. Voluntary-market credits sold to private companies don’t always trigger corresponding adjustments, meaning the host country may still count the reduction in its national inventory while a corporation claims it too.
Which Kinds of Projects Tend to Hold Up
Project type is the single strongest predictor of whether a credit represents real atmospheric benefit. The market divides into avoidance credits, which prevent emissions that would otherwise occur, and removal credits, which pull carbon dioxide out of the atmosphere.
Avoidance credits depend on a counterfactual — you have to estimate what would have happened without the project — and you can never directly observe the emissions that didn’t happen. The uncertainty never fully disappears even in good projects. Removal credits are generally considered more robust because captured carbon can be measured rather than estimated against a hypothetical.
Within those categories, some project types have a much better track record than others.
Methane capture at landfills, dairy operations, and oil and gas wells tends to score well. Methane is roughly 80 times more potent than carbon dioxide over a 20-year period, so preventing its release produces an outsized benefit per ton, and the equipment either exists and captures gas or it doesn’t. Additionality is usually cleaner than in forestry.
Direct air capture with geological storage sits at the high-quality, high-cost end of the market. Purchase prices in 2024 ranged from roughly $100 to $2,000 per ton, averaging near $490 over recent years.7World Resources Institute. 6 Things to Know About Direct Air Capture That is orders of magnitude above nature-based credits, but the tonnage captured is directly measurable and the geological storage is effectively permanent.
Reforestation and avoided deforestation are the cheapest, most popular, and most problematic categories. Avoided-deforestation projects operating under the REDD+ framework pay landowners to preserve forests said to be under threat of clearing.8FAO. REDD+ Reducing Emissions from Deforestation and Forest Degradation The baseline question — would the forest really have been destroyed without the payment? — is hard to answer and easy to inflate. Rigorously verified reforestation on genuinely degraded land can produce real credits, but the category as a whole is where most of the market’s integrity failures have concentrated.
Renewable energy credits in regions where clean power is already economically competitive routinely fail additionality. If the wind farm was going to be built anyway, the credit generates revenue but no additional climate benefit.
How to Tell a Legitimate Credit From a Junk One
A handful of signals separate credits that are likely to represent real reductions from ones that probably don’t.
- Check the registry. Credits should be listed on a recognized platform such as Verra’s Verified Carbon Standard, the Gold Standard, or the American Carbon Registry, with full serialization and public retirement records. Each credit carries a unique serial number that persists through every transfer and retirement, letting anyone trace its history.9Verra. Verified Carbon Standard10ACR Carbon. ACR – World’s First Private Carbon Crediting Program for Offsets
- Look for the CCP label. The Integrity Council for the Voluntary Carbon Market has developed the Core Carbon Principles, ten science-based criteria meant to serve as a global quality benchmark. Credits carrying the CCP label have passed the most rigorous available third-party quality assessment. The council has been assessing major crediting programs against those criteria, and only approved categories from programs that pass are eligible.11ICVCM. The Core Carbon Principles
- Favor removals over avoidance. Removal credits with durable storage are generally more defensible than avoidance credits. Within avoidance, methane capture tends to have cleaner additionality than avoided deforestation.
- Check for corresponding adjustments. For credits from projects in other countries, confirm that the host nation applies a corresponding adjustment under Article 6. Without one, the reduction may be counted twice.
- Read third-party project ratings. Independent rating agencies now evaluate individual projects and issue quality scores. The ratings are imperfect but add analysis beyond what the certifying standard alone provides.
- Check the standard behind the credit. Verra’s VCS is the largest program and also the one at the center of the market’s biggest credibility crisis. The Gold Standard, founded by WWF and other NGOs, requires projects to demonstrate benefits beyond carbon by contributing to at least two UN Sustainable Development Goals. The American Carbon Registry, a nonprofit enterprise of Winrock International, operates in both compliance and voluntary markets with a focus on science-based methodologies. None of these organizations verify projects themselves; they rely on independent auditors accredited under ISO 14065.12SustainCERT. Carbon Credits 101 – What Is the Gold Standard13ACR. ACR Program14Verra. Validation and Verification15ANAB – ANSI National Accreditation Board. ISO 14065 – ANAB
- Be suspicious of the average price. The voluntary market averaged roughly $6 per ton in 2025. A credit at that price is probably worth about what you paid for it.
The Legal Risk of Calling Yourself Carbon Neutral
Buying offsets is one thing. Building a marketing claim on top of them is another, and the legal exposure is growing.
The most prominent U.S. case is a class-action lawsuit against Delta Air Lines, which had marketed itself as “the world’s first carbon-neutral airline.” The complaint alleges that the credits Delta bought from Verra provided no real environmental benefit, making the neutrality claim false. As of late 2025 the case is in active litigation, with Delta opposing class certification. The theory — that offset-based neutrality claims constitute deceptive advertising when the underlying credits are low quality — could shape corporate disclosure practices for years.
Federal guidance has not kept up. The FTC’s Green Guides, last updated in 2012, do include provisions on carbon offset claims.16Federal Trade Commission. Environmentally Friendly Products – FTC’s Green Guides An update was expected to address “carbon neutral” and “net zero” more specifically, but that revision has stalled and appears unlikely under the current administration. In its absence, litigation and state consumer protection laws are filling the gap.
Disclosure requirements are also in limbo. The SEC adopted rules in 2024 that would have required public companies to disclose the costs of carbon offsets used as a material component of their climate targets.17U.S. Securities and Exchange Commission. The Enhancement and Standardization of Climate-Related Disclosures – Final Rules The rules were challenged, stayed pending litigation, and the SEC subsequently voted to withdraw its defense of them.18U.S. Securities and Exchange Commission. SEC Votes to End Defense of Climate Disclosure Rules No federal disclosure requirement for offset spending is currently in force, which leaves corporate offset claims largely self-regulated and correspondingly exposed to private lawsuits.
The Role Offsets Should Play
The mainstream frameworks that evaluate corporate climate claims have converged on a common position: offsets are not a substitute for cutting emissions.
The Science Based Targets Initiative, which certifies corporate climate targets against a 1.5°C pathway, has taken a firm line. The consultation draft of its Corporate Net-Zero Standard Version 2 states that credits may not be “used as a substitute for decarbonizing operations, supply chain activities, or products.”19Science Based Targets. Deep Dive – The Role of Carbon Credits in SBTi Corporate Net-Zero Standard V2 SBTi outlines three narrow use cases where high-quality credits can supplement direct decarbonization: counterbalancing truly residual emissions with removal credits, contributing to mitigation outside a company’s value chain, and making up for underperformance against targets. The language shifts from “offsetting,” which implies cancellation, to “contributing,” which acknowledges that buying a credit doesn’t erase the original emission.
The Oxford Principles for Net Zero Aligned Carbon Offsetting, revised in 2024, lay out the same trajectory as a transition. Organizations should cut emissions first, use high-integrity avoidance credits in the near term for what they cannot yet eliminate, and progressively shift toward removal credits with durable storage as those technologies scale. By the global net-zero target date, any residual emissions should be compensated only with long-lived removals such as direct air capture with geological sequestration.
So: do carbon offsets work? A minority of them do. The rest range from marginal to worthless, and the market has spent 25 years failing to sort them reliably. If you are buying credits, treat them as the last step rather than the first, prioritize removals over avoidance, look for the CCP label and a credible registry, and be careful about what you say publicly. A carbon-neutral claim is only as defensible as the credits underneath it.