How to use RASFF data in a supplier risk review

This guide is for food safety, QA and compliance teams who need to assess an origin or a supplier before approving it, or at an annual review. After reading it, you will be able to pull the EU's own rejection record for an origin in about fifteen minutes, tell a real signal from noise, and set up an alert so you hear about new rejections without checking by hand.

RASFF is the EU's Rapid Alert System for Food and Feed. Every time a member state rejects a consignment at the border, or finds a risk in a product already on the market, it notifies the system. The data is public but hard to read in its raw form. This site reorganises it by origin, product, hazard and week. The steps below use Turkey as the worked example, because it is a large supplier with a mixed record.

Step 1: Start with the origin page

Open the origin's page, for example Turkey. It shows border rejections over the last 12 quarters, a yearly count, and a table by product category.

What to look at: the quarterly trend and the category table. Turkey received 923 border rejections between 2023-Q4 and 2026-Q3. Fruits and vegetables account for 746 of them, and nuts for 93.

Signal vs noise: a single quarter proves little. Turkey's count swings from 40 to 135 per quarter, and in the last two years the peaks came in Q4. Look for a change that lasts more than two quarters, or a new hazard appearing in the "most-cited issues" line.

Pitfall: this page counts border rejections only. RASFF has three main classifications. A border rejection means the consignment was stopped before entering the EU. An alert means a risky product was already on the market and needs rapid action. An information notification is lower urgency. Over the same 12 quarters Turkey also had 138 alerts and 347 information notifications, as well as the 923 rejections. The origin and product pages show rejections. The hazard and weekly pages show every classification. If a supplier's product is already sold in the EU, the alerts matter as much as the rejections.

Step 2: Go down to the product you buy

An origin total mixes products that have nothing in common. Open the product page for what you actually source.

For Turkey, the difference is large. Figs and other tropical fruit had 519 border rejections in the period, mostly for ochratoxin A and aflatoxins. That single product line is more than half of everything Turkey had rejected. Turkish hazelnuts, by contrast, show very few.

What to look at: the product's count, its trend, and the most-cited hazards. The hazards tell you what to test for and what to ask the supplier for in their certificate of analysis.

Step 3: Compare rates, not counts

A large supplier gets more rejections simply because it ships more. Where the data allows it, each product page gives a rate: border rejections per $1 billion of EU imports of that product from that origin, over the last three years. It also ranks the origin against other origins shipping the same product.

Pitfalls:

• Only compare rates within one product. A rate for figs and a rate for hazelnuts are measured against different trade and cannot be ranked against each other. The site never shows a rate for an origin as a whole, for that reason.
• No rate is shown below five rejections. One rejection against a small trade flow produces a huge, meaningless number. A missing rate means "too few cases to say", not "clean".
• The rate is against trade value, not consignments inspected. The EU does not publish how many consignments of each product it inspects, so this is not a failure rate per shipment.

Step 4: Check the EU's own control level

For high-risk products, EU Regulation 2019/1793 sets the share of consignments that must get identity and physical checks at the border. Product pages show it where it applies. Turkish pistachios face 50% checks for aflatoxins, and peppers face 20% for pesticide residues.

Signal vs noise: a check frequency is a policy decision, not an outcome. Groundnuts from Argentina and from the United States both face 20% aflatoxin checks, yet Argentina's rate is 4.7 per $1 billion against 199.0 for the US. A high check level tells you the EU is watching closely, so your supplier's consignments face a higher chance of being held and delayed. It does not tell you the product is failing. Use it for lead-time and cost planning. Use the rate from step 3 for risk.

Step 5: Look at the hazard across all origins

Open the hazard page for your product's main risk, for example aflatoxins. It covers every classification, not just border rejections, and ranks origins. Turkey is the most-cited origin for aflatoxins, followed by the United States and Argentina.

Why it matters: if your alternative supplier is in a different country, this is where you check whether switching reduces the risk or just moves it.

Step 6: Don't confuse who notified with where it came from

Each notification names a notifying country, the member state that found the problem, and one or more origin countries. These are different things. The Bulgaria notifier page shows 1,164 border rejections since 2018, and 1,064 of them are Turkish goods, because much Turkish road freight enters the EU through Bulgaria. Bulgaria notified 317 of Turkey's 923 rejections in the period, 34%.

Use it for: checking the entry point. If most rejections for your product come through one border post and you ship through another, your exposure may differ from the headline. Also, about 4% of notifications name more than one origin. Those count once for each origin they name, so origin totals add up to slightly more than the notification count.

Step 7: Read the latest weeks, carefully

The weekly pages list every notification by week, with origin, product, notifying country and subject. They are the fastest way to see what happened recently.

Pitfall: recent periods are incomplete. The current week and the current quarter are still filling up. On 21 September, week 39 held 18 notifications, while a full week usually holds around 105. The same applies to "this year" totals on every page. Turkey's 248 rejections for 2026 cover less than nine months, not twelve. Never compare an unfinished period with a finished one.

Putting it together

For a supplier review, record five things: the product-level count and trend (step 2), the rate and rank against other origins (step 3), the EU check frequency (step 4), the main hazards (steps 2 and 5), and any recent alerts on products already on the market (step 1). That is enough to justify a testing plan, a certificate-of-analysis requirement, or a decision to qualify a second origin.

Set up a watch

Checking these pages every week by hand is how reviews go stale. Set up a watch for the origin you buy from, and you will get at most one email a week, sent when new notifications match it.