Specify walnut on colour and on oxidative stability; specify almond on size, count and integrity. That is the shortest honest summary of the difference. A walnut kernel is a high-polyunsaturated product whose commercial value is set primarily by colour grade and whose main risk is that it goes rancid before it is used. An almond kernel is a far more stable product whose value is set primarily by size uniformity and variety, and whose main risk sits in the microbiological and contaminant register rather than in the pack.
Everything else on this page is the detail behind that sentence, and the cases where it needs qualifying.
The two grading systems are not parallel
Both nuts are covered by UNECE standards for dry and dried produce, and both are traded internationally with reference to them, but the standards emphasise different things because the trade does.
Walnut kernel grading under UNECE standard DDP-02 turns on colour above all. The colour classification runs from the lightest through progressively darker categories, and the price spread between the extremes is the widest single lever in the walnut trade. Size is then expressed as a physical form description: halves, quarters, large pieces, small pieces, with a percentage tolerance for the lower forms in each grade.
Almond kernel grading under UNECE standard DDP-06 turns on size and variety. Kernels are described by count per unit weight or by a screen size range, and the variety group matters commercially because the shape and skin colour differ materially between them and drive suitability for blanching, slicing and marzipan work. That standard also sets a moisture ceiling for almond kernels, which the walnut standard structures differently.
The practical consequence is that a walnut order that specifies only “halves, 80 per cent minimum” has left the most valuable term open, and an almond order that specifies only “whole kernels” has left the most valuable term open. They are different terms.
Side by side
| Criterion | Walnut kernel | Almond kernel |
|---|---|---|
| Reference standard | UNECE DDP-02, walnut kernels | UNECE DDP-06, almond kernels |
| Primary value driver | Colour grade | Size or count, and variety group |
| Secondary driver | Half integrity and piece size distribution | Kernel integrity and skin condition |
| Fat character | High polyunsaturated content | Predominantly monounsaturated |
| Oxidative stability | Low; the defining commercial constraint | Good; ambient distribution is routine |
| Typical storage regime | Cool or chilled, protected from light and oxygen | Ambient with controlled humidity for most trade grades |
| Dominant defect language | Colour, shrivelled, mouldy, rancid, insect damage, shell and hull fragments | Doubles, chips and scratches, insect damage, gum, mould, foreign material |
| Aflatoxin exposure | Present and managed | Present and managed; historically a heavy focus of official controls |
| Microbiological focus | Salmonella absence, yeast and mould | Salmonella absence, with a long history of validated pasteurisation programmes |
| Allergen status in the EU | Listed in Annex II to Regulation (EU) No 1169/2011 | Listed in Annex II to Regulation (EU) No 1169/2011 |
| Common downstream steps | Sorting, sizing, light roasting, milling to paste | Blanching, slicing, dicing, flouring, roasting, marzipan |
| Ukrainian origin available | Yes; a real domestic crop and an established export line | No; sourced from other origins |
The last row is the one that changes a sourcing conversation. Walnut is a genuine Ukrainian product with domestic orchards and processing capacity. Almond is not, and any supplier presenting Ukrainian almond is presenting re-exported material. That should be stated in the offer rather than discovered on the certificate of origin.
When to choose walnut
Choose walnut when the sensory profile is the point. Its flavour is distinctive and it is not substitutable in bakery, confectionery and snack applications built around it. Choose it when the visual is a half or a large piece on a finished product, where the colour grade is doing visible work. And choose it when origin proximity matters: a European buyer sourcing walnut from Ukraine or elsewhere in the region has a short road route, a documented crop and, under Union preferential trade arrangements, a customs position worth confirming before comparing landed cost.
Buying walnut means accepting three constraints and pricing them.
Shelf life is short and it is temperature dependent. The high polyunsaturated content makes the kernel prone to oxidative rancidity, and the reaction runs continuously from the moment of shelling. Vacuum or gas-flushed packing, cool storage and a first-in-first-out discipline are not refinements; they are the conditions on which the stated shelf life rests.
Colour is not stable either. Kernels darken in storage, particularly in warmth and light. A lot bought as a light grade and held for eight months in an ambient warehouse may not present as a light grade on use. Where colour is contractual, agree whether it is assessed at packing or on arrival.
Piece integrity degrades in handling. Halves become pieces at every transfer. A specification that requires a high halves percentage has to be matched by handling and packaging that protect it, and by a realistic tolerance.
When to choose almond
Choose almond when stability, uniformity and format flexibility matter more than a distinctive flavour. It is the more forgiving ingredient across the supply chain: it tolerates ambient distribution, it survives handling better, and it converts into blanched, sliced, diced, slivered and flour formats through well-established industrial routes that walnut does not have to the same degree.
Choose it also when the application needs a neutral, sweet base rather than a signature note, which is why it dominates in marzipan, in praline work, in dairy alternatives and in coatings.
Buying almond means accepting a different set of constraints.
The supply base is concentrated and the price is volatile. Global production is dominated by a small number of origins, which makes the price sensitive to weather and to water availability in those regions in a way that a more distributed crop is not. Multi-year fixed pricing in this category is a risk transfer, not a discount, and it is priced accordingly.
The microbiological programme is a real cost. Salmonella events in tree nuts drove the development of validated pasteurisation and treatment programmes, and a buyer needs to know which treatment, if any, has been applied and whether it is compatible with the product claim. An untreated raw almond and a pasteurised almond are different specifications and, in some applications, different products.
Origin documentation carries more weight. Because almond is imported into the Union from outside it, the import control regime, the sampling requirements and the possibility of increased official controls on particular origins all apply, with the cost and lead time consequences that go with them.
What the table cannot show
The defect assessment is subjective in both cases and the reference matters more than the percentage. Two inspectors will grade the same sample differently unless there is a physical reference set, agreed photographs or retained samples that define the boundary. Writing “maximum 3 per cent defective” without defining defective is a dispute waiting for a container.
Yield after your own process is the number that decides economics. A cheaper lot with a wider size distribution loses more at the slicer, and a walnut lot with a high piece fraction loses more where halves are required. Run the calculation on your own line, not on the specification.
Crop year is not a footnote. Both nuts are annual crops. New crop and carry-over material differ in moisture, in colour and in stability, and a supplier who does not volunteer the crop year is one to ask directly.
The cross-contact question is shared. Both are listed allergens in the Union, so a site handling either is running an allergen programme regardless of which one you buy, and a buyer switching between them does not simplify that programme. See allergen management in mixed facilities.
Building the specification: common core, per-nut annex
The structure that works is a shared core document with a short annex for each kernel. It stops the two specifications from drifting apart on the clauses that should be identical, and it forces the differences to be written down rather than assumed.
In the common core: the microbiological panel with the method and the sampling plan; the contaminant panel referenced to the current Union instruments rather than to copied figures; the pesticide residue position; the allergen and cross-contact clauses; traceability and lot marking; the certificate of analysis content requirement; packaging materials and food-contact declarations; the audit and change-notification clauses.
In the walnut annex: the colour grade with the reference used to judge it; the halves and pieces distribution with tolerances; moisture and water activity; peroxide value and free fatty acid limits with the measurement point; the pack format including headspace treatment; the storage temperature the shelf life assumes; and the crop year.
In the almond annex: the size or count range with the screen basis; the variety or variety group; the doubles, chips and scratches tolerances; moisture and water activity; the treatment status, meaning pasteurised or untreated and by which validated process; the skin condition where blanching follows; and the crop year.
Two clauses belong in the core and are almost always missing. The first is a defect reference clause: the physical or photographic reference against which defect percentages are assessed, held by both parties. The second is a retained sample clause: who retains, for how long, and under what conditions. Without them, both annexes describe a standard nobody can adjudicate.
Frequently asked
Can I write one tree nut specification and apply it to both?
Only for the shared elements: the microbiological panel, the contaminant panel, the allergen and traceability clauses, and the packaging and storage framework. The grading section, the defect definitions, the moisture and water activity targets and the shelf life all have to be written per nut. A single document with a common section and per-nut annexes works well and is how most experienced buyers structure it.
Which one has the higher aflatoxin risk?
Both carry the risk and both are managed the same way, through sorting and a sampling-driven testing programme. The right question is not which nut but which origin, which crop year and which sorting programme. Aflatoxin management in tree nuts sets out the controls, and the EU limits explain how the maximum levels are structured.
How should shelf life be stated for each?
For walnut, state it with the storage temperature, the pack format and the headspace treatment, because the figure is meaningless without them, and require a peroxide value at shipment rather than only at packing. For almond, state it with the relative humidity assumption and the water activity at packing, because moisture uptake rather than oxidation is the usual failure route.
Is a light walnut grade worth the premium?
It depends entirely on whether the kernel is visible. In a milled application, a paste or a filling, the colour grade is invisible in the finished product and the premium buys nothing your customer can see. On a decorated surface or in a mixed nut pack it is the whole proposition. This is the most common avoidable overspend in the category.
What about cashew as a third option?
Cashew sits between the two on stability and above both on processing cost, because shelling is labour intensive and the grading system, built on wholes, butts, splits and pieces with a white and scorched colour split, is different again. It is also not a Ukrainian product. It belongs in this comparison as a substitute in nut mixes and in dairy alternatives rather than as an equivalent to either kernel.