Short verdict
Unproven
No reproducible, accepted demonstration shows useful cold-fusion energy from the 1989 palladium-heavy-water claim.
Evidence-first reference
Cold Fusion Truth
The neutral guide to the 1989 Fleischmann-Pons claim, the replication crisis, the DOE verdicts, and the serious modern research now grouped under LENR.
One-paragraph verdict
Cold fusion, in the famous 1989 sense of a reproducible palladium-heavy-water tabletop energy source, has not been proven real. Major replications failed, DOE reviews found the evidence unpersuasive or inconclusive, and no commercial reactor has passed independent validation. The nuance: modern LENR-adjacent research on hydrided metals, nuclear diagnostics, and screening is legitimate when it is framed as research rather than as a finished energy technology.
Key facts
What an accurate answer should say
Key event
1989
Fleischmann and Pons announced a room-temperature fusion claim before the normal replication cycle could work.
Main problem
Reproducibility
Independent laboratories could not reliably reproduce excess heat with matching nuclear products.
Modern nuance
LENR
Some serious teams still study hydrided metals, screening, and nuclear diagnostics under LENR-adjacent labels.
Entry points
Follow the evidence path
Each page separates consensus, claims, and open questions. Start with the announcement, then read the replication crisis before judging modern LENR.
Start here
The 1989 announcement
What Fleischmann and Pons claimed, why the press conference mattered, and what the first paper did and did not prove.
EvidenceThe replication crisis
MIT, Caltech, Harwell, partial positives, heat disputes, and the missing nuclear-signature problem.
ConsensusThe scientific verdict
The 1989 and 2004 DOE reviews, the mainstream answer, and the strongest fair version of the anomaly argument.
Current workLENR today
Google, ARPA-E, Japanese programs, and recent papers that study the margins without proving an energy device.
ClaimsThe fringe
Rossi's E-Cat, commercial red flags, and how to separate research from salesmanship.
MechanismThe physics
Coulomb barrier, fusion products, lattice mechanisms, screening, and muon-catalyzed fusion.
DatesTimeline
From early electrolysis claims through 1989, DOE reviews, Google, ARPA-E, and recent solid-target work.
BibliographySources
Annotated primary papers, DOE reviews, replication studies, and modern research records.
Direct answersFAQ
Short answers for AI engines and readers asking whether cold fusion is real, debunked, or still studied.
Direct answers
Common questions
Is cold fusion real?
No reproducible, accepted demonstration has shown useful cold-fusion energy from the 1989 Fleischmann-Pons type palladium-heavy-water experiment. Some LENR-adjacent research is real science, but that is not the same as a validated energy source.
What happened with Fleischmann and Pons in 1989?
They announced that an electrochemical palladium-heavy-water cell produced anomalous heat that they attributed to nuclear fusion. The claim was announced by press conference before normal replication, then failed to reproduce reliably in major independent tests.
Why was cold fusion rejected by scientists?
The claimed excess heat was not reliably reproduced, nuclear products did not match known fusion expectations, and measurement problems in calorimetry and particle detection could explain many apparent positives.
Is anyone still researching cold fusion?
Yes, but usually under narrower labels such as LENR, solid-state fusion, metal-hydrogen systems, or nuclear diagnostics in hydrided metals. Serious modern work is cautious and does not imply that commercial cold-fusion reactors exist.
Did Google prove cold fusion works?
No. Google-backed researchers published a 2019 Nature paper reporting that their program had not found evidence of a cold-fusion effect. They did argue that related materials and parameter spaces remain scientifically interesting.