One-Up / One-Down Traceability (USDA + Global)

This topic is part of the SG Systems Global traceability, regulatory & operations glossary.

Updated November 2025 • One-step-forward / one-step-back, USDA, FDA, EU, GFSI, end-to-end genealogy, EPCIS, WMS/MES integration • Operations, Quality, Regulatory, Supply Chain, IT/OT

One-up / one-down traceability (also called “one step forward / one step back”) is the minimum regulatory expectation that every company in the food chain knows from whom it received product (one step back) and to whom it passed it on (one step forward). Under USDA, FDA, EU and other regimes, this is the baseline requirement for recalls and incident investigations: no one expects you to magically know what your supplier’s supplier did, but they absolutely expect you to know exactly which lots you bought from whom and which customers you shipped them to. Globally, it’s the most common denominator of traceability. It is also, bluntly, not enough for modern risk, branding and efficiency expectations.

“One-up / one-down is the regulatory floor, not the ceiling. If that’s all you have, your recalls will be wide, your answers will be slow, and your ‘digital traceability’ will fall apart the moment someone asks hard questions.”

1) The Core Concept: One Step Back, One Step Forward

The idea is simple enough that regulators, lawyers and auditors can all recite it:

• One step back – for any product you hold or have transformed, you must be able to identify which supplier(s) and lot(s) it came from.
• One step forward – for any lot you held or shipped, you must know which customer(s) and locations received it.

In a linear supply chain, linking every node this way produces an overall trace from farm to fork, at least in theory. Each company is responsible for its own link: from last supplier in to first customer out. That’s the baseline regulators use to reconstruct the chain during incidents. If you can’t do even that, you’re non-compliant before the conversation really starts.

2) One-Up / One-Down Under USDA & US Law

In the U.S., the one-up / one-down concept appears across agencies:

• USDA FSIS – expects official establishments to keep distribution and receiving records sufficient to trace product lots to immediate suppliers and consignees. FSIS directives and guidance are very clear about batch/lot identifiers, production dates and shipping detail expectations.
• FDA / FSMA – under FSMA, “one-up/one-back” recordkeeping remains a core expectation for most foods, with additional traceability lot code requirements for high-risk foods under the Food Traceability Rule.
• Retail & distribution – often underpins their own internal recall and incident procedures and is embedded in vendor agreements.

The regulatory language varies, but the practical expectation does not: you must be able to show FSIS or FDA, quickly, which lots you got from whom and where they went. “We think it was these trucks” is not acceptable. Detailed purchase, receiving, production and shipping records are the evidence; one-up/one-down is the relationship those records must support.

3) Global Context: EU, Codex & GFSI

Globally, the one-up / one-down idea shows up repeatedly:

• EU General Food Law – requires food and feed business operators to be able to identify their immediate suppliers and customers (with some exceptions for direct-to-consumer). Same principle: one step back, one step forward.
• Codex Alimentarius – recommends that food business operators maintain information to identify from whom and to whom food has been supplied.
• GFSI schemes (BRCGS, SQF, FSSC 22000, etc.) – all require traceability systems that at minimum support one-up/one-down and can be tested via mock recalls.

Different jurisdictions add their own twists—record retention times, specific data elements, “key data elements” and “critical tracking events”—but the minimum is always the same. If you export or sell into multiple markets, designing once for strong one-up/one-down and then adding local layers on top is smarter than reinventing the basics for each customer or regulator.

4) What One-Up / One-Down Is Not

There are plenty of misconceptions worth killing off:

• It is not full internal genealogy – regulators are satisfied if you can identify lots and shipments; they do not require you to track every internal movement or transformation (even though you should, for your own benefit).
• It is not a complete mass-balance model – one-up/one-down doesn’t require you to reconcile every kilogram; that’s an internal yield and fraud-control choice.
• It is not a real-time consumer-facing traceability app – nice for marketing, but regulators are still focused on your ability to produce accurate records in hours, not QR-code storytelling.
• It is not enough on its own for complex multi-ingredient, multi-site operations – it tells you who you bought from and who you shipped to, but not necessarily how lots were blended across dozens of processes inside your own four walls.

One-up/one-down is table stakes for compliance. End-to-end traceability inside your facility and across your group is what allows you to manage risk and money properly. Confusing the two is how companies end up technically compliant yet operationally blind during incidents.

5) Basic Data Elements: What You Actually Need to Know

To do one-up/one-down properly, you must capture and keep, at minimum:

• Supplier linkage – supplier identity, purchase orders, delivery dates, lot/batch codes, quantities, product codes (internal & supplier), program attributes (organic, halal, etc.).
• Internal lot ID – the lot or batch code you assign on intake or first transformation (ideally linked to GS1-128 intake labeling and SSCCs).
• Production linkage – for transformed products, a mapping between input lots and output lots, at least at “family” or batch level.
• Customer linkage – customer IDs, invoices, shipment records, SSCCs or pallet IDs, ship dates, destination DCs/stores, and the lots/pallets/cases included.

If any of these links relies on hand-written notes, ad-hoc spreadsheets or “we can reconstruct it if we have time”, your one-up/one-down capability is fragile. It may scrape through a friendly mock recall; it will break under the pressure of a real outbreak investigation where regulators expect complete answers quickly.

6) GS1, SSCC & EPCIS: Standardising One-Up / One-Down

Global trade does not work well if everyone uses their own idea of “lot ID.” That’s where GS1 standards come in:

• GTIN – standard product identifier used in GS1-128 labels for cases and pallets.
• Lot/batch (AI 10) – encoded in GS1-128 and used as your internal and external lot key.
• SSCC (AI 00) – unique pallet/logistics unit ID, crucial for lot transfer scanning and shipments.
• EPCIS – standard for sharing event-based traceability data (receiving, shipping, transformation, aggregation) across organisations.

When one-up/one-down is built on GS1 and EPCIS, “who we got it from” and “who we sent it to” become machine-readable events rather than PDF attachments and CSV uploads. That’s the direction regulators and major retailers are pushing: standardised, near-real-time one-up/one-down, not hand-curated spreadsheets after the fact.

7) One-Up / One-Down vs Internal End-to-End Genealogy

Inside your own facility (or group of plants), you can do far more than one-up/one-down. Internal genealogy covers:

• Every transformation event (grind, mix, cook, slice, pack) with explicit input–output relationships.
• Every lot transfer event between locations and statuses (as in GS1-128 Lot Transfer Scanning).
• Mass balance across intake, WIP and finished, exposing yield, shrink and unexplained loss.
• Program segregation across lines and sites (organic, halal, retailer programs).

External one-up/one-down uses a thin slice of that internal data: just enough to say “we got these lots from Supplier X and shipped these lots to Customer Y.” This is why modern plants treat one-up/one-down as part of the output of their internal genealogy engine, not as a separate system. You don’t want to maintain two versions of the truth; you want rich internal data that can be sliced into the simple one-up/one-down story whenever regulators or customers ask for it.

8) Mock Recalls: Proving One-Up / One-Down Works

Every GFSI scheme and most customer codes of practice require regular mock recalls. These tests are where one-up/one-down gets exposed:

• You are given a suspect lot from a supplier and asked to list all finished products and customers affected.
• Or you are given a finished SKU and asked to trace back to contributing raw lots and suppliers.
• Or you are asked to trace both directions (back and forward) within a fixed, short time window (often 2–4 hours).

If your systems support true one-up/one-down, this is painful but doable: run a query, pull a report, sanity-check, send. If they don’t, mock recalls become all-hands-on-deck spreadsheet sessions, and the findings are politely ignored until the next audit. That is the red flag: if mock recall output requires heroics every time, your one-up/one-down will collapse in a real crisis when the clock and the lawyers are both running faster than you can search PDFs.

9) Where It Goes Wrong: “We Can Reconstruct It If We Have Time”

Common failure patterns in one-up/one-down implementations:

• Non-unique lots – lot codes reused by suppliers or internally across too much time, making it impossible to isolate events to a narrow window.
• Missing or inconsistent IDs – hand-written tags, whiteboard notes, “temporary” IDs that don’t match system records.
• Unscanned transfers – pallets and totes moved without GS1-128 scans, leaving location and status unknown or wrong.
• Blended records – multiple raw lots recorded as a single generic lot in production; multiple outbound shipments lumped into aggregate reports without SSCC-level detail.
• Paper islands – key traceability links (e.g., blast chill logs, smokehouse load cards) exist only on paper outside MES/WMS, with no reliable cross-reference to lot IDs.

All of these make one-up/one-down brittle. The system might look fine for straightforward inbound/outbound flows; the moment regulators ask about a specific product date, line or DC, the edges fray. The cure is brutal but simple: design processes so that nothing can move or transform without the right scan and ID, and then remove paper as a primary traceability medium everywhere possible.

10) Linking One-Up / One-Down to FSMA, FSIS & Retailer Demands

Regulators and retailers are raising the bar beyond simplistic one-up/one-down:

• FSMA Food Traceability Rule – for listed high-risk foods, requires more granular key data elements and critical tracking events than classic one-up/one-down, often across multiple nodes.
• FSIS expectations – in meat/poultry, FSIS increasingly expects plants to produce traceability data that covers more than one step when internal commingling is complex.
• Retail customer codes – often demand lot-level traceability to DC and sometimes store level, with expectations of mock recall completion in hours, not days.

One-up/one-down remains the legal baseline, but customer expectations and specific rules are pushing the industry towards more detailed, digital traceability. Designing your system to do the minimum today is designing to be obsolete tomorrow. It’s smarter to start from an end-to-end mindset and treat one-up/one-down as an easy derived view rather than the core objective.

11) Implementation Roadmap: From Paper to Digital One-Up / One-Down

A practical route from minimal to solid one-up/one-down capability:

• Map the chain – identify all inbound and outbound nodes: which suppliers, which receiving points, which customers, which shipping points.
• Standardise IDs – adopt consistent lot/batch ID rules; introduce GS1-128 with GTIN/lot/SSCC where appropriate; stop accepting ambiguous or duplicated IDs.
• Digitise key links – implement scan-based receiving, lot transfer scanning for internal moves, and SSCC-level shipment capture.
• Integrate WMS/MES/ERP – ensure lot IDs and SSCCs flow cleanly across systems; avoid duplicate entry and conflicting records.
• Test with mock recalls – run progressively more challenging scenarios; fix the gaps; repeat until the exercise stops being painful.

Don’t try to “blockchain the supply chain” before you can reliably say which lots you shipped yesterday and where. Solid one-up/one-down, backed by GS1 and decent system integration, is the foundation for all the fancier stuff vendors are currently hyping.

12) FAQ

Q1. Is one-up / one-down traceability really the legal requirement everywhere?
It is the core principle in most major jurisdictions (US, EU, many others): each operator must know its immediate supplier(s) and customer(s) for food lots. Specific rules (e.g., FSMA traceability, EU additional recordkeeping) may demand more detail for certain foods, but one-up/one-down is the baseline logic almost everywhere.

Q2. If I can trace from supplier to finished product inside my plant, is that more than one-up/one-down?
Yes. That’s internal end-to-end genealogy. One-up/one-down would only require that you link suppliers to intake lots and intake lots to outbound lots/shipments. The richer genealogy is for your own benefit—tighter recall scope, better yield, program integrity—not because regulators insist on every internal detail (yet).

Q3. Do I need GS1-128 and SSCCs to meet one-up/one-down expectations?
Strictly speaking, no; any consistent ID system could work on paper. Practically, GS1-128 and SSCCs massively reduce friction, especially when dealing with multiple plants, 3PLs and retailers. They also open the door to EPCIS-based data sharing later. Designing without GS1 now usually creates extra work and mapping headaches later.

Q4. How fast should I be able to respond to a regulator or customer traceability query?
Regulations may speak in hours or days; customers increasingly expect answers in hours at worst, sometimes minutes if you’ve automated well. If you need more than a working day to answer a focused “where did this lot go?” or “what fed this product?” question, your one-up/one-down system is not really under control.

Q5. Is one-up / one-down enough for brand protection?
Barely. It will get you through the minimum regulatory bar for recalls, but it won’t optimise recall scope, yield, or support nuanced program claims. For serious brand protection you need deeper internal genealogy, mass balance, and, ideally, the ability to trace not just who but how much and under which conditions—one-up/one-down is just one lens on that richer dataset.

Related Reading
• Identification & Standards: GS1-128 Case Label | SSCC | GS1 Application Identifiers (AIs)
• Traceability & Events: End-to-End Lot Genealogy | GS1-128 Lot Transfer Scanning | Intake-to-Grind Digital Handover | EPCIS Traceability Standard
• Programs & Compliance: Mock Recall Performance | Food Safety Plan (FSP) | HACCP | Data Integrity