A supplier hands you a PDF called "LM-79 Test Report." The numbers look good. But three things might be true: the report is for a different product, the lab doesn't exist, or the data was typed into Excel, not measured in a sphere. Here's how to tell the difference — before you wire $18,000.
LED test reports come in a three-letter alphabet soup. Suppliers throw around "LM-79 tested" and "L70 > 50,000 hours" like they're interchangeable. They're not. Here's what each standard actually measures — and what it can't tell you.
| Standard | What It Measures | What It Can't Tell You | Test Duration |
|---|---|---|---|
| LM-79 | Complete fixture performance: total lumens, efficacy (lm/W), CCT, CRI, chromaticity, power factor | How performance changes over time. An LM-79 is a single-point measurement. | 2–4 hours |
| LM-80 | LED package/chip lumen maintenance at 3 case temperatures (55°C, 85°C, and a third manufacturer-selected point) | Fixture-level performance. LM-80 tests only the LED chip on a heat sink in a controlled chamber — not the complete fixture. | 6,000–12,000+ hours |
| TM-21 | Projected L70 lifetime — the hours until light output drops to 70% of initial | Actual end-of-life. TM-21 is a mathematical projection, limited to 6× the LM-80 test duration. | Calculation only |
Here's the thing: a supplier who shows you an LM-79 is doing the bare minimum. One who also provides LM-80 data from the LED chip manufacturer AND a TM-21 projection that connects the two — that's a supplier who understands their supply chain.
Counterfeit test reports are more common than you'd think. We've verified 3,200+ product listings on Compare2Best, and roughly 12% of supplier-provided test documents had at least one verifiability problem — expired accreditation, mismatched product identifiers, or report numbers that didn't trace back to any lab.
Every legitimate LM-79 report carries the testing laboratory's accreditation mark. Look for these logos in the report header or footer:
| Accreditation Body | Region | What to Look For |
|---|---|---|
| NVLAP (National Voluntary Laboratory Accreditation Program) | USA | NVLAP Lab Code, e.g. "NVLAP Lab Code 200901-0" |
| CNAS (China National Accreditation Service) | China | CNAS registration number, e.g. "CNAS L1234" |
| ILAC-MRA | Global mutual recognition | ILAC-MRA mark alongside the national accreditation body's logo |
| TAF (Taiwan Accreditation Foundation) | Taiwan | TAF accreditation number |
| KOLAS (Korea Laboratory Accreditation Scheme) | Korea | KOLAS accreditation number |
If the report has no accreditation logo at all, or only the supplier's own company logo, treat it as a marketing document — not test data. ISO/IEC 17025 accreditation is the minimum bar for photometric testing laboratories.
Every accredited lab issues a unique report number. Most major labs (UL, Intertek, TÜV, SGS, Bureau Veritas, DEKRA) have online verification portals. If the lab exists but doesn't have a public portal, call or email them with the report number and ask: "Is this your report, and does the product description match?"
We've seen suppliers change one digit in a real report number — same format, different report. The lab confirms: "That report belongs to a different client and a different product."
A real LM-79 report states these parameters explicitly:
A report missing two or more of these isn't necessarily fake, but it's not a professional test either. Professional labs document conditions because reproducibility matters.
Real integrating sphere measurements have minor asymmetries. Luminous intensity distribution curves show small irregularities. CCT values drift slightly between measurement positions.
If the photometric data is perfectly symmetrical — every angle showing exactly the predicted value — you're looking at a spreadsheet, not a sphere. This is surprisingly common: the supplier takes the LED chip datasheet numbers, plugs them into an Excel IES file generator, and saves it as a "test report." The IES file will load in lighting design software, but it represents the chip's theoretical performance, not the fixture's actual output.
| Order Value | Verification Required | Cost | Time |
|---|---|---|---|
| < $2,000 | Verify report number with lab; confirm accreditation logo | Free–$0 | 10 min |
| $2,000–$10,000 | Above + request LM-80 data from LED chip manufacturer + TM-21 projection | Free (supplier should provide) | 1–3 days |
| $10,000–$50,000 | Above + commission independent LM-79 spot test on a random production sample | $300–800 | 5–10 days |
| > $50,000 | Above + ISTMT for thermal validation + full batch audit | $1,500–4,000 | 2–4 weeks |
When a supplier says "L70 > 50,000 hours," here's where that number comes from — and where it can be manipulated.
TM-21 is the projection standard. It takes LM-80 data (real measurements of lumen depreciation over time) and fits an exponential decay curve to project when output will hit 70%. The math has strict limits: you can only project 6× the test duration. So a 6,000-hour LM-80 test can project to 36,000 hours max. An L70 claim of 50,000 hours requires at least 8,334 hours of LM-80 data.
Suppliers who claim L70 > 100,000 hours with only 6,000 hours of LM-80 data are violating the standard's projection limit. TM-21 explicitly prohibits projections beyond 6× the test duration.
| LM-80 Test Duration | Max TM-21 Projection | What This Covers |
|---|---|---|
| 6,000 hours | 36,000 hours | ~4 years continuous operation |
| 10,000 hours | 60,000 hours | ~7 years continuous operation |
| 12,000 hours | 72,000 hours | ~8 years continuous operation |
| 15,000 hours | 90,000 hours | ~10 years continuous operation |
Not every LED application needs every test. Match your verification to your risk.
| Application | Must-Have Reports | Nice-to-Have |
|---|---|---|
| Office/commercial indoor | LM-79 (lumens, CCT, CRI, efficacy) | LM-80 + TM-21 |
| Warehouse/industrial high-bay | LM-79 + TM-21 | ISTMT (thermal) |
| Outdoor/street lighting | LM-79 + TM-21 + ISTMT | LM-82 (temperature operating point) |
| DLC Premium qualification | LM-79 + LM-80 + TM-21 + ISTMT + LM-82 | LM-84 (lumen & color maintenance) |
| Horticultural lighting | LM-79 + PPF/PPFD (IES LM-79-19 Annex) | Spectral power distribution |
| Retail/display lighting | LM-79 (CCT, CRI, chromaticity) | TM-30-18 (color fidelity) |
| Hazardous location (Class I/II) | LM-79 + UL 844 / ATEX certification | TM-21 |
For orders where trust matters more than the test fee, commission independent testing. Major photometric labs operate globally:
| Lab | LM-79 Cost (Approx.) | Turnaround | Accreditation |
|---|---|---|---|
| UL (USA) | $500–900 | 7–10 business days | NVLAP |
| Intertek (global) | $400–800 | 5–8 business days | NVLAP, CNAS, UKAS |
| SGS (global) | $350–700 | 7–12 business days | CNAS, NVLAP |
| TÜV Rheinland | €300–600 | 5–10 business days | DAkkS |
| DEKRA | $400–750 | 5–8 business days | NVLAP, CNAS |
| Bureau Veritas | $350–650 | 7–10 business days | CNAS, COFRAC |
Process: request the supplier ship one random production sample directly to the lab (not to you, not to a middleman). Tell the supplier you're doing a spot check — their reaction tells you as much as the test results. A supplier who says "sure, which lab?" is confident. One who makes excuses about "proprietary technology" or "special version for testing" is telling you something.
No. LM-80 measures the LED package alone under controlled thermal conditions. Your fixture's thermal management — heat sink design, driver placement, enclosure airflow — determines how hot the LEDs actually run. The same LED chip in a well-cooled fixture hits L70 at 60,000 hours; in a poorly-designed one, it hits L70 at 18,000 hours. Always request TM-21 data calculated for the fixture's actual case temperature, not the chip manufacturer's ideal value.
An IES file contains photometric data but no accreditation trace. A real IES file generated from an LM-79 test will have the lab's name and report number embedded in the header (visible when opened in a text editor). If the IES file header shows only the manufacturer's name, it's likely generated from a spec sheet — not measured. Always get the full LM-79 PDF report alongside the IES file. The PDF has the accreditation stamps and test conditions the IES doesn't carry.
Test reports don't formally expire, but their relevance decays with production changes. If the supplier changes the LED chip, driver, lens, or housing material, the original LM-79 is no longer valid for the new version. DLC requires re-testing every 5 years even for unchanged products. Industry best practice: if the report is older than 3 years, ask the supplier to confirm in writing that the current production matches the tested sample — and verify with a spot check.
LM-79 uses CRI (Ra) — the average of the first 8 color samples (R1–R8). TM-30-18 is a newer standard that measures 99 color samples and reports fidelity (Rf) and gamut (Rg). For most commercial lighting, CRI ≥ 80 or ≥ 90 from LM-79 is sufficient. For retail, museum, or healthcare where color accuracy is critical, request TM-30 data. CRI can be gamed — a fixture can score Ra 90 while rendering reds (R9) at −20. TM-30 makes that impossible to hide.
Compare verified LED products with real test reports
Every product on Compare2Best links to its supporting documentation — LM-79, certifications, and supplier verification status.
Compare LED Products on Lighting Hub
This guide is produced by the Compare2Best knowledge team and reviewed by lighting industry experts.
Published: July 13, 2026 · Reading time: ~9 min