Nuclear data evaluatorSalary, qualifications, career path and hiring demand, 2026 edition
A nuclear data evaluator converts measurements, nuclear-reaction theory, nuclear engineering requirements, and benchmark evidence into the evaluated datasets used by reactor physics, criticality safety, shielding, fusion, safeguards, isotope production, and radiation transport codes. The job sits between experiment and application: compiling EXFOR evidence, fitting resonance parameters, running reaction models, estimating covariances, formatting ENDF/GNDS files, and validating them against integral benchmarks. Unlike a reactor physicist, the evaluator’s product is the nuclear data itself rather than one reactor calculation built from that data. They also apply advanced technologies and data analytics to coherently present complex materials, ensuring the nation's critical infrastructure benefits from innovative nuclear energy solutions.
TRX models established US nuclear data evaluators at roughly $130,000–$160,000 base, senior evaluators at $155,000–$190,000 and principal specialists around $180,000–$215,000. The broader May 2025 BLS physicist median is $172,250. In the UK, established evaluators model around £56,000–£70,000, rising into £80,000+ for principal and programme roles because the talent pool is extremely small.
A PhD in nuclear physics, nuclear engineering or a closely related field is the normal gate. Employers want evidence that the candidate can combine experimental datasets, nuclear models, statistics and transport-code feedback into a defensible evaluation. ENDF-6/GNDS formats, EXFOR, resonance analysis, TALYS/EMPIRE/CoH or equivalent reaction modelling, covariance methods, NJOY/PREPRO processing and benchmark validation are recurring filters.
The role at a glance
everything an employer will ask about in the first fifteen minutes of a screening call.
- Also called
- Nuclear data scientist · nuclear reaction data evaluator · evaluated data physicist · cross-section evaluator · nuclear data analyst · ENDF evaluator · data visualization tools degree · computer science · software tools
- Entry qualification
- PhD in nuclear physics, nuclear engineering, applied physics or related technical discipline is standard. Exceptional MSc candidates can enter through processing, benchmarking or database roles and grow into evaluation. Data science with Python and machine learning concepts are increasingly valuable.
- Typical entry pay
- $105,000–$130,000 US TRX model · £46,000–£58,000 UK.
- Senior pay
- $155,000–$190,000 US · £68,000–£84,000 UK, with principal and international programme leadership extending above those ranges.
- Contract day rates
- £550–£800/day experienced evaluator · $80–$120/hr US specialist support, with niche fission-yield, covariance or resonance work potentially higher.
- Professional gate
- No PE/CEng requirement. Peer-reviewed nuclear data work, evaluator authorship, recognised library contributions and benchmark-validation credibility matter more. Compliance with internationally accepted professional standards and government agencies’ regulations is essential.
- Security
- Civil library work is generally open science; national-security, naval, safeguards or weapons-related programmes may require citizenship, federal background investigation, selective service system registration, and security clearance. Periodic drug testing and pre employment drug screen are common.
- Where the work sits
- BNL NNDC, LANL, ORNL, LLNL, IAEA Nuclear Data Section, OECD-NEA Data Bank, CEA, JRC, UKNNL, JAEA and specialist university groups. Idaho National Laboratory is a key US hub.
- Travel
- Moderate. CSEWG, JEFF Nuclear Data Week, IAEA technical meetings, WPEC/INDEN work and collaborator visits create regular international travel.
- Shift pattern
- Normal research hours. Library-release deadlines, benchmark campaigns and international working groups create peak periods rather than shift work.
- TRX segments
- Reactor physics · Criticality safety · Fusion · Fuel cycle · Nuclear medicine · Safeguards · Scientific computing · data driven analysis · systems performance analysis
six versions of the same job title
nuclear data evaluation covers different physical quantities and applications. A resonance evaluator, fission-yield specialist and thermal scattering law evaluator use different physics, but all produce traceable evaluated data for downstream calculations.
Neutron reaction cross-section evaluator
Evaluates elastic, inelastic, capture, fission and other neutron-induced reaction data across resonance and fast-energy regions using measurements, nuclear theory and integral feedback. This is the core ENDF/JEFF evaluator profile.
Resolved / unresolved resonance evaluator
Fits resonance energies, widths, spin/parity assignments and covariance information using transmission, capture and fission measurements. R-matrix methods, SAMMY or equivalent tools and uncertainty treatment are central.
Fission yield / decay data evaluator
Produces independent/cumulative fission yields, decay schemes, half-lives, branching and uncertainty data used in decay heat, safeguards, isotope inventories and reactor analysis.
Thermal scattering law evaluator
Builds S(α,β) evaluations for moderators and materials where molecular/solid-state effects change low-energy neutron scattering. Graphite, light/heavy water, hydrides and advanced moderators are typical systems.
Covariance / uncertainty evaluator
Quantifies uncertainties and correlations in nuclear data using Bayesian, generalized least-squares and model-based methods. The output feeds sensitivity, uncertainty and adjustment studies in criticality and reactor design.
Library validation / integral data evaluator
Tests candidate evaluations against criticality, shielding, reactor physics and other integral benchmarks, tracing biases back to specific reactions or energy regions before library release.
What the week actually looks like
a composite day for a senior evaluator preparing a structural-material neutron evaluation for a future library release.
What nuclear data evaluators are paid in 2026
“Nuclear data evaluator” is too small a profession for an official national salary series. The ladder below is a TRX market model anchored to physicist and nuclear-engineer pay plus the premium associated with national-laboratory research, specialised coding and international library authority.
How nuclear data evaluation compares to adjacent roles
National-laboratory evaluator pay varies by grade and clearance. The TRX ladder reflects the scarcity of evaluator-level expertise rather than assuming every nuclear physicist earns the same premium.
| Occupation | Median | P10 | P90 | What moves the number |
|---|---|---|---|---|
| Nuclear data evaluator | $160,000 | $105,000 | $215,000 | Library authorship, isotope ownership, covariance, benchmark validation and programme authority |
| Physicists | $172,250 | $82,110 | $274,110 | Official BLS May 2025 benchmark |
| Nuclear engineers | ~$134,000 | ~$93,000 | ~$196,000 | Official BLS May 2025 engineering benchmark |
| Physical scientists, all other | $129,230 mean | — | — | Broad federal/research fallback comparator |
National-laboratory evaluator pay varies by grade and clearance. The TRX ladder reflects the scarcity of evaluator-level expertise rather than assuming every nuclear physicist earns the same premium.
Major evaluated isotope ownership
Authoring or leading widely used evaluations for uranium, plutonium, iron, oxygen or other high-impact nuclides carries strong technical authority.
Covariance / Bayesian evaluation
Reliable uncertainty quantification remains harder to recruit than pointwise cross-section work alone.
Evaluation + application validation
Evaluators who can move from EXFOR/model fitting into criticality, shielding and reactor benchmarks are more valuable than single-domain specialists.
Three ways in
The dominant route is nuclear physics PhD → reaction modelling or experiment → formal evaluation. A second route comes through reactor physics and benchmark analysis, but candidates still need enough reaction physics to own data rather than merely consume it.
Nuclear physics PhD route
Reactor physics / criticality route
Data / computational physics route
Are you actually ready to compete for a nuclear data evaluator role?
A strong CV names the nuclides, reactions, energy range, datasets, models, evaluation format, covariance method, processing tools and benchmarks you personally owned. “Worked with ENDF data” is a user statement. “Re-evaluated Fe-56 inelastic scattering using EXFOR, model constraints and SINBAD/criticality validation” is evaluator evidence.
Free resume scoring on avua. Your score is yours; it is not shared with employers.The biggest uplift comes from proving ownership of a formal evaluation or covariance/validation package rather than only publishing measurements or model calculations.
Illustrative TRX shortlisting pattern only.
The credentials that actually gate the work
Nuclear data evaluation has no professional licence. Credibility is built through advanced physics, recognised evaluation methods and traceable contributions to libraries or international working groups.
| Credential | Jurisdiction | Required for | Time | Notes |
|---|---|---|---|---|
| PhD in nuclear physics / nuclear engineering | Global | Standard evaluator entry | 4–6 yrs | Reaction physics, fission, resonance or computational nuclear science are common. |
| ENDF-6 / GNDS format competence | US / international | Library production | Months–years | ENDF/B-VIII.1 is distributed in ENDF-6 and GNDS formats. |
| EXFOR / experimental data competence | International | Evidence compilation | Experience-based | Evaluator must assess original measurements, metadata and uncertainties. |
| Nuclear reaction / R-matrix modelling | Reaction-specific | Evaluation generation | Years | TALYS, EMPIRE, CoH, SAMMY or equivalent toolchains. |
| Covariance / Bayesian methods | Senior roles | Uncertainty evaluation | Years | Increasingly central to JEFF, ENDF and application uncertainty work. |
| Processing / transport validation | Library work | Release qualification | Experience-based | NJOY, PREPRO, AMPX plus MCNP/SCALE/OpenMC or equivalents. |
| Peer-review / library contribution record | Senior evaluator | Technical authority | Multi-year | CSEWG, JEFF, INDEN, WPEC and evaluated-file authorship are strong signals. |
| Security clearance | Programme-specific | National-security data | Months | Not required for open civil libraries; selected DOE/defence work differs. |
ENDF/B-VIII.1 is now released in both ENDF-6 and GNDS, and NNDC explicitly encourages users and developers to test GNDS workflows ahead of future library evolution. New evaluators increasingly need both nuclear physics and modern data-structure/software fluency.
What appears on a 2026 nuclear data evaluator shortlist
Employers screen for someone who can defend every stage of the chain from raw measurement to transport-code result.
Named on the specification
- Nuclear reaction / decay physics — reaction mechanisms, resonances, fission, angular distributions, spectra, conservation laws and decay/fission-yield systematics.
- Experimental data assessment — EXFOR use, original publication review, normalisation, systematic uncertainty, correlation and dataset consistency.
- Evaluation modelling — R-matrix, optical model, Hauser-Feshbach, pre-equilibrium or other appropriate nuclear-model methods.
- ENDF / GNDS file production and verification — formal data structures, checking codes, conservation and internal consistency.
- Covariance / uncertainty quantification — statistically defensible uncertainty and correlation information tied to measurements and models.
- Processing and benchmark validation — NJOY/PREPRO/AMPX plus criticality, shielding or reactor benchmarks used to test application impact.
What decides between two shortlisted candidates
- High-impact actinide evaluation — uranium/plutonium or other fission-system nuclides with direct reactor and criticality relevance.
- Fission yield / decay heat expertise — scarce capability spanning safeguards, decay heat and inventory applications.
- Thermal scattering law evaluation — specialist bridge between condensed-matter physics and neutron transport.
- Automated evaluation pipelines — reproducible workflows, optimisation, CI/testing and GNDS-native tooling.
- Machine learning / Bayesian inference — emerging evaluation and uncertainty methods with traceability and physics constraints.
- International library leadership — visible roles in CSEWG, JEFF, INDEN or WPEC and coordination across laboratories/countries.
The 2026 demand map
Nuclear data is a very small labour market, but 2026 has unusually visible library and methods activity: ENDF/B-VIII.1 publication, JEFF-4 development, GNDS adoption, covariance work and advanced reactor/fusion data needs.
| Programme | Location | Phase in 2026 | Engineering demand |
|---|---|---|---|
| NNDC / CSEWG ENDF/B | Brookhaven + US labs | ENDF/B-VIII.1 released and maintained | Very high specialist relevance — 558 evaluations, errata, processing and future ENDF work |
| LANL / US nuclear data programmes | New Mexico, US | Library processing, evaluation and validation | Very high — strong actinide, transport and national-security demand |
| IAEA INDEN | Vienna / international | Active coordinated evaluation projects in 2026 | Very high — structural materials, actinides and internationally adopted evaluations |
| OECD-NEA JEFF | Europe / NEA countries | JEFF-4.0/4.1 roadmap under 2025–2028 mandate | Very high — explicit priority to build evaluation skills and engage new evaluators |
| APRENDE / JEFF Nuclear Data Week | Europe | Active 2026 benchmarking/evaluation programme | High / current — links measurement, evaluation, benchmark and application needs |
| UKNNL nuclear/reaction physics | UK | Applied nuclear physics and reactor support | High niche relevance — domestic reactor, medical isotope, safeguards and decommissioning data needs |
| Advanced reactor / molten-salt data needs | US / Europe | Design and licensing growth | Growing — chlorine, fluorine, salt, fast-spectrum and covariance priorities |
| Fusion nuclear data programmes | International | DEMO/STEP and materials qualification | Growing — charged-particle, activation, transmutation and high-energy neutron data needs |
Nuclear data is a very small labour market, but 2026 has unusually visible library and methods activity: ENDF/B-VIII.1 publication, JEFF-4 development, GNDS adoption, covariance work and advanced reactor/fusion data needs.
the work is concentrated in a handful of institutions.
Most nuclear companies consume libraries; only a small number of national laboratories, international organisations and specialist university groups create them. That means vacancy volume is low but replacement difficulty is high. The JEFF 2025–2028 mandate explicitly calls for encouraging new evaluators because library continuity depends on retaining specialist knowledge that can take a decade to build.
data structure and uncertainty are becoming first-class problems.
ENDF/B-VIII.1 now ships in both ENDF-6 and GNDS, and the NEA’s GNDS programme is preparing the ecosystem for data that legacy ENDF-6 cannot represent cleanly. At the same time, IAEA 2026 meetings focus on uncertainty quantification, Bayesian methods and AI/ML. Evaluators who combine physics with modern software and reproducible inference are therefore increasingly valuable. The practical implication for hiring is that “nuclear data” is no longer only a legacy-format physics role. Teams increasingly need people who can make an evaluation reproducible: version-controlled source data, scripted model runs, automated format checks, covariance propagation, processed-library generation and benchmark regression testing. Candidates who can preserve evaluator judgement while making the workflow auditable and repeatable are especially well placed for future ENDF, JEFF and GNDS-based programmes. That combination also shortens peer review and release cycles because another evaluator can reproduce the data selection, model assumptions, processing steps and benchmark consequences rather than reverse-engineering a one-off analysis from notebooks and informal files.
Adjacent and onward roles
Nuclear data evaluators usually progress into principal scientific authority, library leadership, reactor-physics application leadership or international nuclear-data programmes.
Questions candidates genuinely ask recruiters
How much does a nuclear data evaluator earn in 2026?
TRX models established US nuclear data evaluators at roughly $130,000–$160,000 base, senior evaluators at $155,000–$190,000 and principal specialists at $180,000–$215,000. The May 2025 BLS physicist median is $172,250. In the UK, established nuclear data evaluators model around £56,000–£70,000, with principal and programme roles around £80,000–£115,000 depending on nuclear domain expertise and employer.
What does a nuclear data evaluator actually do?
They combine experimental measurements, nuclear-reaction theory, nuclear engineering requirements and benchmark evidence to produce recommended nuclear data. Typical outputs include neutron cross sections, resonance parameters, angular distributions, emitted-particle spectra, fission yields, decay data, covariances and data visualization tools. The evaluator documents data choices, builds an ENDF/GNDS file, processes it with data analysis tools and tests how it performs in transport, criticality, shielding or nuclear systems calculations.
What is the difference between a nuclear data evaluator and a reactor physicist?
The evaluator creates or improves the underlying nuclear data library using physics informed machine learning and artificial intelligence. The reactor physicist uses that library to calculate a specific reactor, criticality or shielding system. Reactor-physics feedback is essential to evaluation because benchmark biases can expose data problems, but the evaluator must return to measurement, nuclear engineering domain knowledge and reaction physics before changing the recommended data.
Do you need a PhD?
Usually yes. The role requires enough nuclear-reaction physics, statistics, computational data science team collaboration and computational depth to defend an evaluation independently. PhDs in nuclear physics, nuclear engineering context and computational physics are the most common routes, with MSc-level candidates able to enter through nuclear data processing, databases or benchmark analysis and later grow into evaluator responsibility.
Which nuclear data libraries matter most in 2026?
ENDF/B-VIII.1 is the current recommended US library. JEFF is the major collaborative European/NEA library programme, with JEFF-4.0 and follow-on work central to the 2025–2028 mandate. JENDL, CENDL, TENDL and other national and international governments' libraries remain important. Evaluators also work heavily with EXFOR experimental data, GNDS as a next-generation data structure and data sensor fusion methodologies.
What is the most valuable experience for a senior nuclear data evaluator?
Owning a complete evaluation that survives independent review and application testing. Employers want the isotope/reaction, experimental evidence, model, covariance approach, ENDF/GNDS implementation, processing, benchmark results and resulting library contribution. High-impact actinides, structural materials, fission data, nuclear simulation codes and physics informed machine learning carry particular weight because errors propagate into many critical infrastructure and nuclear power plants applications.
We only recruit in nuclear. That is the whole point.
TRX can assess whether your background fits nuclear data evaluation, reaction modelling, covariance/UQ, reactor physics, criticality validation or scientific computing. If you come from experimental nuclear physics, Bayesian inference, transport codes or nuclear databases, we can identify which experience transfers directly into evaluator roles and where formal library-production depth is still missing.