Guide
Raw & Freeze-Dried Pet Food: the Pathogen-Control Operating Manual
Raw, freeze-dried and frozen pet food sits under the strictest pathogen-control bar in the category. FDA Compliance Policy Guide 690.800 sets zero tolerance for Salmonella in finished raw pet food — any positive is a recall. Listeria, STEC and Campylobacter follow the same enforcement posture in practice. The category's growth has run alongside a steady recall cadence, and FDA inspections of raw producers are among the most intensive in animal food. This guide is the operating manual for raw and freeze-dried pet food pathogen control under 21 CFR 507 and CPG 690.800.
Start free trial Free trial, no credit card, onboard in days, not months.
CPG 690.800 — the zero-tolerance frame
FDA Compliance Policy Guide Sec. 690.800 'Salmonella in Food for Animals' establishes that any pet food (including raw and frozen) testing positive for Salmonella is considered adulterated. The CPG is enforcement policy, not a regulation, but is the binding inspection standard. Listeria monocytogenes, Shiga-toxin-producing E. coli and Campylobacter are not in CPG 690.800 explicitly but are treated as recall-trigger pathogens in raw pet food by current FDA enforcement practice. This is the most aggressive enforcement posture in the animal-food space.
The validated kill step — HPP, freeze-dry, test-and-hold
Three operational models dominate. (1) High-Pressure Processing (HPP): typically 600 MPa for 3+ minutes, validated against Salmonella, Listeria and STEC surrogates — the dominant kill step in commercial raw pet food. (2) Freeze-drying with a documented kill component: freeze-drying alone is not a reliable kill step (it preserves pathogens) so most freeze-dried raw uses HPP pre-dry or a separate kill validation. (3) Test-and-hold: hold every lot pending Salmonella negative result — operationally heavy and recall-vulnerable on false negatives, used mainly by small producers.
HPP validation — what FDA expects to see
Validation evidence: a challenge study (in-house or contract lab) demonstrating ≥5-log reduction of Salmonella, Listeria and STEC surrogates at the chosen pressure × time × product matrix; verification of pressure-time profile on every cycle; product-matrix bracketing (poultry, beef, lamb, pork, mixed) so each matrix has supporting kill data; cold-spot mapping in pouch and cycle. HPP equipment qualification (IQ/OQ/PQ) is foundational. Inadequate matrix bracketing is the most common HPP finding — a single beef validation doesn't cover a new lamb formula.
Incoming raw materials — the dominant hazard source
Salmonella in raw pet food originates almost entirely from incoming raw meat, poultry and organs. Supplier qualification is non-optional: USDA-inspected suppliers with documented Salmonella performance, COA per lot, periodic onsite or third-party audits, and a clear non-conforming material disposition. Many producers source from HACCP-certified human-grade processors specifically to reduce incoming Salmonella loading. Cold-chain integrity from supplier through HPP is part of the hazard analysis — temperature abuse pre-HPP increases pathogen load beyond the validated kill capability.
Finished-product testing — sampling plan and lot disposition
Sampling plans for raw pet food are aggressive: typical protocol takes 30 × 25g (or higher) per lot for Salmonella by FDA BAM, plus Listeria and STEC on a defined frequency. Lots are held pending negative result (test-and-hold) or released on the HPP cycle record with retrospective testing as verification. A finished-product positive triggers immediate recall of the lot plus any in-process or finished lots that may share an exposure pathway — the recall scope tends to be broad because the kill step is the only control.
Labelling — handling instructions and the human-illness vector
Raw pet food labels carry explicit handling instructions: keep frozen until use, thaw in refrigerator, use within X days of thawing, wash hands and surfaces after contact, do not feed to immunocompromised humans by mistake. AAFCO Model Regulation language and CDC/FDA recommendations both apply. The label is part of the food-safety control — the courts and FDA have held that inadequate handling instructions contribute to misbranding when human illness traces back to handling exposure.
A 90-day raw / freeze-dried pet-food readiness path
Days 1–15: CPG 690.800 gap assessment and recall-decision authority defined. Days 16–35: kill-step validation review — HPP matrix bracketing, freeze-dry kill component, test-and-hold plan. Days 36–55: incoming raw-material supplier programme refresh — USDA inspection, COA, cold chain. Days 56–70: finished-product sampling plan rebuild and lot-disposition workflow. Days 71–80: label handling-instruction audit. Days 81–85: mock FDA CVM raw pet-food inspection. Days 86–90: management review and CAPA closure.
Where this lives in V5 Ultimate
The clauses above aren't theoretical — every one maps to a shipped module and an industry profile. Jump to the parts of the product that turn this guide into evidence on a Monday morning.
Modules that do the work
HPP validation lifecycle
Matrix bracketing, surrogate studies and ongoing pressure-time monitoring.
CPG 690.800 recall gate
Auto-routed recall-decision authority on any finished-product Salmonella positive.
Handling-instruction control
Raw pet-food label wording verified against current CDC/FDA guidance.
Industries this hits hardest
Frequently asked
Is freeze-drying a kill step for Salmonella?
No — freeze-drying preserves pathogens rather than killing them. A freeze-dried raw product needs either an HPP step before freeze-drying, a validated thermal pre-treatment, or a rigorous test-and-hold programme on finished product. Marketing language about freeze-drying 'destroying' pathogens is not supported by validation literature and FDA inspectors push back on it.
Does HPP validation transfer across product matrices?
Only with bracketing evidence. HPP kill kinetics differ by matrix composition (fat, protein, water activity), pouch geometry and cycle parameters. A validation on a beef matrix doesn't automatically cover lamb, poultry or mixed-protein products. Inadequate matrix bracketing is the most common HPP-related FDA finding — validate the matrices you actually run, with bracketing studies justifying the inference between them.
Can I rely on test-and-hold without a validated kill step?
Operationally yes, but the risk profile is high. Salmonella sampling plans have known false-negative rates, holding cost is significant, and a single missed lot can be a regulatory and brand catastrophe. Most commercial-scale raw pet-food producers run HPP plus finished-product testing because HPP shifts the kill from a sampling-statistics control to a process control. Test-and-hold-only is viable mainly at small scale with very aggressive sampling.
What's the difference between CPG 690.800 and a 21 CFR 507 preventive control?
CPG 690.800 is FDA enforcement policy declaring Salmonella-positive animal food adulterated. 21 CFR 507 is the FSMA regulation requiring a written food-safety plan with preventive controls for hazards including Salmonella. They layer: Part 507 is the system you build (hazard analysis, HPP validation, EMP, supplier programme), CPG 690.800 is the enforcement standard against which finished product is judged. You need both.
See it on your shop floor.
Free trial, no credit card, onboard in days, not months.
Related reading
Spot something off? .