Hazard class tells you what kind of danger you are dealing with. Hazardous goods categories (the packing groups) tell you how much of it. That second number is the one that decides what box you legally have to use, how much you can put in it, and whether an aircraft will take it.
Hazardous goods categories are also the part of classification people most often guess at, because they look like a judgement call. They are not. Every packing group is derived from a measured property against a published threshold. This guide sets out those thresholds class by class, and shows what each one costs you downstream.
What hazardous goods categories actually are
Three groups, applied within a hazard class:
- Packing group I — high danger
- Packing group II — medium danger
- Packing group III — low danger
The group is not a property of the class; it is a property of the substance within that class. A Class 3 flammable liquid can be PG I, II or III depending on how readily it ignites. Two products with the same UN number can even differ, where an entry admits more than one group.
Some classes have no groups at all. Classes 1, 2 and 7 never take one, and lithium batteries in Class 9 do not either. Writing "2.1, PG II" on a shipping paper is a straightforward error, and a common one.
How it is decided: flammable liquids
Class 3 uses two measured numbers — flash point and initial boiling point.
| Packing group | Flash point (closed cup) | Initial boiling point | Typical product |
|---|---|---|---|
| I | Any | ≤ 35 °C | Diethyl ether, some fuel additives |
| II | < 23 °C | > 35 °C | Acetone, ethanol, most perfume bases |
| III | ≥ 23 °C and ≤ 60 °C | > 35 °C | White spirit, some diffuser oils |
Notice where the line falls. 23 °C separates PG II from PG III, and that is close enough to a warm British afternoon that small reformulations can move a product across it. A perfume base at 21 °C flash point is PG II. Adjust the ethanol content, land at 25 °C, and it becomes PG III with materially easier packaging and quantity limits.
Corrosives, where concentration rules
Class 8 uses skin destruction time and metal corrosion rate:
- PG I — destroys full-thickness skin within 3 minutes, after an exposure of up to 3 minutes
- PG II — destroys skin within 60 minutes, after exposure of up to 1 hour
- PG III — destroys skin within 4 hours, or corrodes steel or aluminium faster than 6.25 mm per year at 55 °C
Because these are all dose-dependent, concentration decides the answer. Sodium hydroxide at 30 % is a different packing group from the same chemical at 2 %, and at low enough strength it may not be regulated at all. This is why a Class 8 classification made against an ingredient rather than the finished mixture is nearly always wrong. For a full worked example of concentration deciding the group, see our guide to sodium hypochlorite dangerous goods classification, which walks through household bleach against industrial strength.
Toxics, measured in LD50
Division 6.1 uses toxicity data (oral LD50, dermal LD50, inhalation LC50) with thresholds set out in 49 CFR 173.132. The most severe route governs. A substance with moderate oral toxicity but high inhalation toxicity is graded on the inhalation figure.
Why the group hits your costs, not just your paperwork
This is the part people underestimate. The packing group sets the UN performance packaging standard your outer packaging must meet, and those standards are tested and certified.
| Packing group | Packaging mark | Drop test height | Relative air quantity limit |
|---|---|---|---|
| I | X — approved for PG I, II and III | 1.8 m | Smallest |
| II | Y — approved for PG II and III | 1.2 m | Moderate |
| III | Z — approved for PG III only | 0.8 m | Largest |
The practical read: a box marked Y can carry PG II and PG III but not PG I. If your classification says PG II and your packaging is Z-rated, the shipment is non-compliant before it leaves the building — and nothing on the label will tell the picker that.
A reformulation that nobody flagged
Take a candle and home fragrance brand. Their reed diffuser had shipped as UN1993, Class 3, packing group III for two years, in Z-rated bottles inside a Z-rated outer. Perfectly correct.
Then procurement changed solvent supplier to shave cost. The new base carried more isopropanol, and the flash point dropped from about 26 °C to about 19 °C. Crossing 23 °C moved the product from packing group III to packing group II.
Nothing visible changed. Same bottle, same label, same listing, same SKU. The packaging was now under-specified, the air quantity limits were tighter than the pack size allowed, and the shipping papers had been wrong since the first pallet of new stock arrived. It surfaced during a carrier audit, which is the expensive way to find out.
A packing group is a measurement, not a property of the product name. When the formulation moves, the group can move with it — and the only reliable signal is re-running the classification against the new numbers.
Classify7 ruleset team, on why the engine asks for a flash point rather than assuming one
How hazardous goods categories change what you can send
Packaging standard is the obvious consequence. The less obvious one, and often the more painful commercially, is quantity.
Every entry carries maximum net quantities per package, and those limits tighten sharply as the packing group rises. They tighten again between cargo aircraft and passenger aircraft. The effect is that a packing group I liquid can be technically shippable and commercially useless by air, because the permitted quantity per package is smaller than a saleable unit.
This is why hazardous goods categories deserve attention at product development rather than at despatch. A formulation decision that moves a product from packing group II to packing group III can:
- Increase the permitted quantity per package, sometimes substantially
- Allow a cheaper Z-rated outer instead of a Y-rated one
- Open passenger aircraft services that were previously closed
- Bring the product inside limited quantity relief, removing a large amount of paperwork
None of that is visible from a marketing brief. It surfaces only when someone runs the classification against the measured properties — which is an argument for doing it while a formulation is still changeable, not after the first production run.
The quantity limitation columns sit in the Hazardous Materials Table at 49 CFR 172.101, and the air figures are in the corresponding IATA DGR entry.
When the answer requires a lab
Sometimes there is no way around it. If your SDS gives a flash point range rather than a value, or gives an open-cup figure where the criteria want closed-cup, or predates a reformulation, then the honest position is that you do not yet know the packing group.
The correct response is to get the number, not to pick the middle option. Classify7 handles this by asking for the measured property where the group depends on one, and returning needs_review rather than inventing a group when the property is missing. An unanswered field is recoverable. A confidently wrong one propagates onto every shipping paper you generate.
- Assigning a packing group to Class 2, Class 7 or a lithium battery entry, none of which take one.
- Using an open-cup flash point against closed-cup criteria.
- Classifying the neat ingredient rather than the diluted mixture you actually ship.
- Keeping Z-rated packaging after a classification moved to packing group II.
- Never re-running the classification after a supplier or recipe change.
Where this fits
Packing group is the fourth field in a complete determination, after UN number, proper shipping name and hazard class. For the classes themselves, see the nine classes of hazardous materials. For finding the entry in the first place, use the UN number lookup reference. And to read the flash point off a supplier document correctly, our guide to reading an SDS for classification shows which sections to trust.
The full criteria live in 49 CFR part 173 on eCFR, free to read and worth bookmarking.
