Does FDA require raw fish to be frozen before serving as sushi or sashimi?

Yes for most species. FDA 21 CFR 123 (HACCP Chapter 5) requires either −20°C for 7 days or −35°C for 15 hours to destroy Anisakis and other parasites. Key exemptions: tuna species (low natural Anisakis burden) and aquaculture fish raised on formulated feed with no ocean prey exposure — the infection pathway structurally cannot exist.

Is raw fish safe to eat without being frozen?

It depends on the fish's origin. Wild-caught fish requires freezing to destroy Anisakis parasites. Farm-raised fish on formulated feed — like Sasshu Salmon and Goto Islands Bluefin Tuna — qualify for a freezing exemption under FDA Food Code 3-402.11(B)(2) because the parasite lifecycle cannot complete in a closed, controlled system. Sashimi DC's products are documented per shipment.

What Is Anisakis — and How Does It Reach Your Fish

Anisakis is a genus of parasitic nematodes (roundworms) found throughout the world's oceans. The species responsible for most human infections is Anisakis simplex sensu stricto — accounting for 88.4% of molecularly identified anisakiasis cases in Japan (Sugiyama et al., 2022). A second species, A. pegreffii, is the predominant cause in Europe and Korea. Pseudoterranova azarasi also causes human infections, particularly in northern Japan (Hokkaido) among consumers of Pacific cod.

The lifecycle is the key to understanding both the risk and how it is managed:

The Anisakis lifecycle

From whale to plate — and how aquaculture breaks the chain

In wild fish (risk present)

Adult worms live in the stomach of marine mammals (whales, dolphins, seals). Eggs pass into seawater via feces. Free-living larvae hatch and are consumed by copepods and krill (first intermediate hosts). Fish eat the infected crustaceans; larvae migrate into fish muscle and viscera. Marine mammals eat the fish, completing the cycle. Humans are accidental dead-end hosts.

In formulated-feed aquaculture (no risk)

Pelleted feed is heated and extruded during manufacture — any viable parasite material is destroyed. Fish raised exclusively on pellets never eat infected krill or small fish. The first step of the transmission chain — ingestion of infected prey — cannot occur. Anisakis cannot enter the fish. This is not risk reduction; it is structural elimination of the hazard.

Studies confirm: a large-scale survey of more than 4,000 farmed Norwegian salmon found zero anisakids (Levsen et al., 2019). The risk ratio for farmed vs. wild Atlantic salmon is approximately 570:1 lower in farmed fish (Eurosurveillance, 2021). A Danish survey of farmed rainbow trout (350 fish, freshwater and marine-cultured) similarly found no zoonotic parasites (Lymbery et al., 2021) — wild cod sampled by the same method carried five nematode species, confirming the method's reliability.

How Common Is Anisakiasis

Japan is the world's leading case nation for anisakiasis — a direct consequence of high raw seafood consumption in the form of sushi and sashimi. A landmark 2022 study published in CDC Emerging Infectious Diseases (Sugiyama et al.) used health insurance claims data from 2018–2019 to estimate the true annual incidence: approximately 19,737 cases per year — an average of 21,511 in 2018 and 17,962 in 2019. Official food poisoning statistics reported only <1/40th of that number, indicating massive underreporting. Most infections are gastric (larvae in the stomach wall, causing severe pain), but intestinal anisakiasis — more difficult to diagnose — accounts for a significant share. Approximately 13% of endoscopically confirmed cases in the study were asymptomatic.

In the United States, anisakiasis is recognized but far less common — partly because American seafood is more likely to have been frozen (for distribution logistics reasons as much as for safety), and partly because raw fish consumption per capita remains lower than in Japan. However, as sushi consumption has grown in the US, cases have increased. The American Gastroenterological Association recommended preventive controls for parasite-containing species intended for raw consumption, which underlies the FDA HACCP requirement.

All 23 asymptomatic cases (13% of the 177 confirmed patients) were discovered incidentally during cancer screening or routine health check-ups — meaning a significant number of people carry Anisakis larvae without any symptoms. Asymptomatic cases were nonetheless confirmed via endoscopy. This underlines that the incidence figures, while striking, still undercount actual prevalence.

Which Fish Carry the Most Anisakis Risk

Anisakis burden varies enormously by species and origin. The critical variable is not geography but feeding ecology: how much time the fish spends eating infected marine prey (euphausiids, copepods, small pelagic fish) and how long it has accumulated parasites. The research picture is now quite clear:

Anisakis risk by species — wild fish

Prevalence data from Japanese waters and Alaska (peer-reviewed)

Highest risk (wild)

Chum salmon (O. keta): 100% prevalence in Japanese waters and near-universally in Alaska. Wild Atlantic salmon: 100% in Scottish and Norwegian coastal studies. Blue mackerel (S. australasicus): 97.5%. Alaska pollock: 100%. Pacific saury (sanma): high prevalence, one of the leading causes of anisakiasis in Japan. Wild pink salmon: high and increasing. Herring: approaches 100% in North Sea populations. Hake (European): 76.7% in Portuguese samples.

Lower risk (wild) / Zero risk (farmed closed-tank)

Bluefin tuna: notably lower Anisakis burden than the species above — high trophic level predator; larvae from infected prey rarely establish at high density in tuna viscera. Sasshu Salmon (Sashimi DC): zero risk. Raised in a closed land-based tank in Kagoshima on formulated feed with no ocean access. The infection pathway structurally cannot exist — no euphausiid or wild-prey exposure at any point. This is not a claim about farmed salmon as a category; it is specific to the controlled, documented production system at Satsuma Sendai Unagi.

The species with the highest case burden in Japan — mackerel (saba), Pacific saury (sanma), and squid (ika) — are eaten raw in traditional preparations (shiime, ikizukuri) that do not involve freezing. Wild salmon, the highest-prevalence species in Alaska, is covered by the FDA HACCP freezing requirement specifically because of its documented muscle-level larval burden.

Wild Salmon — Why the Fillet Itself Is the Risk

With most fish, the common assumption is that parasite risk is concentrated in the viscera — and evisceration (gutting) takes care of it. Wild Pacific salmon breaks that assumption. A 2024 longitudinal study published in PLOS ONE examined archived canned fillets of four Alaskan salmon species (chum, pink, sockeye, coho) processed between 1979 and 2019. The finding: more than 90% of anisakid larvae were recovered from edible muscle tissue — belly flap, loins, and tail meat — not just the visceral cavity.

This is not a post-mortem migration artifact. It reflects genuine intra-vitam muscle penetration that occurs during the ocean phase. Wild Pacific salmon spend years feeding intensively at sea on euphausiids and infected forage fish, accumulating larvae that migrate through the visceral wall into the surrounding musculature while the fish is still alive. The belly flap (hypaxial muscle, anatomically adjacent to the body cavity) consistently shows the highest concentration. Dorsal muscle and tail sections carry larvae too.

The practical implication: for wild Pacific salmon, gutting at sea or at the dock removes the visible visceral mass but does not remove the larvae already encysted in the flesh. The fillet you buy has already accumulated its parasite burden. This is precisely why FDA HACCP requires freezing for wild salmon sold for raw consumption — and why the formulated-feed exemption matters: Sasshu Salmon's burden is zero at the source, not reduced by processing.

A related finding from Scotland and Norway: wild Atlantic salmon larval intensity increases with sea age and body weight — older, heavier fish carry proportionally more larvae. A fish caught at the river mouth on its spawning return retains its full ocean-accumulated burden even though it has stopped feeding. Parasite load does not decline when the fish leaves saltwater.

The seasonal and geographic picture. Anisakiasis cases in Japan peak in October–December (coinciding with the autumn mackerel and squid season) and again in April–June (spring saury and herring). The US and Europe consume the same infected wild species but largely frozen or cooked — which is why Japan accounts for an estimated 90%+ of global reported anisakiasis cases despite not having unusually high parasite densities in its fish. The key variable is preparation, not geography. One genuine biology trend: Anisakis intensity in wild Atlantic salmon has increased approximately fourfold since 2009 in Scottish coastal waters, driven by recovering grey seal populations following marine mammal protection legislation. More marine mammals = more definitive hosts shedding eggs = higher parasite pressure on prey fish. This is a conservation success story with a food safety footnote.

Hotaruika and the Crassicauda Nematode

Firefly squid deserve a separate treatment because the parasite situation is specific and different from the standard Anisakis story. The nematode larvae found in hotaruika viscera belong to the genus Crassicauda — specifically larvae whose adult form (C. giliakiana and related species) parasitizes beaked whales. These larvae accumulate in the viscera of Watasenia scintillans as an intermediate host en route to their whale definitive host. Molecular characterization of these larvae has confirmed their identity and the whale-squid transmission pathway (Goto et al., 2023, International Journal for Parasitology: Parasites and Wildlife).

The risk to humans comes from eating the raw viscera of infected squid. Symptoms can include: severe abdominal pain, nausea, diarrhea, and in more serious cases, cutaneous larva migrans (creeping eruption under the skin) and intestinal obstruction requiring medical intervention. Historical Japanese medical case series documented this pattern. Recent surveillance data from the Japan Institute for Health Security (IASR Vol. 46, January 2025; Kumagai et al., 2023) places the infection rate at 0.6–1.6% in recent surveys — down from 2–7% in pre-2011 studies, a decline attributed to reduced Mesoplodon beaked whale populations in the Sea of Japan.

In Japan, the standard response has always been to boil the squid immediately at the port after landing — not primarily as a food safety measure, but because boiling is the preferred preparation for texture and flavor. The parasite-control effect is incidental but complete: the heat destroys nematode larvae. The result is what Sashimi DC carries — lightly boiled hotaruika that is both the safest and the most flavorful form of the ingredient.

For consumers who want to eat raw hotaruika: whole-body deep freezing per FDA protocol is the correct control (not just marinating or removing the visible viscera). Sashimi of the visceral-free mantle only is also safe.

Note on terminology: Many general sources use "Anisakis" as a catch-all term for parasite risk in raw fish. Technically, hotaruika's parasite is Crassicauda, not Anisakis — a different nematode genus with a different lifecycle. Both are destroyed by proper freezing or cooking. Sashimi DC's own hotaruika page uses the correct terminology (see the FAQ below), but the distinction is worth knowing when researching the topic elsewhere.

FDA HACCP — The Freezing Requirements (21 CFR 123)

The FDA governs parasite control in seafood under the Hazard Analysis and Critical Control Points (HACCP) regulation, codified at 21 CFR Part 123. Chapter 5 of the Fish and Fishery Products Hazards and Controls Guidance covers parasites in detail. The core requirement: processors must treat parasites as a significant hazard whenever they know or have reason to know the fish will be consumed without thorough cooking.

Three equivalent freezing protocols are recognized — any one is sufficient:

1

−20°C for 7 Days

Standard long-freeze protocol

Freeze and store at an ambient temperature of −4°F (−20°C) or below for a minimum of 168 hours (7 days total). This is the minimum-temperature protocol and is the most commonly used in commercial distribution because it is achievable with standard commercial freezers. Note that −18°C (the typical home freezer setting) does not meet the −20°C threshold.

2

−35°C Until Solid, then −35°C for 15 Hours

Rapid-freeze protocol A

Freeze at an ambient temperature of −31°F (−35°C) or below until solid, then store at −31°F (−35°C) or below for a minimum of 15 hours. This achieves parasite destruction much faster than protocol ①. The higher temperature accelerates freezing throughout the product.

3

−35°C Until Solid, then −20°C for 24 Hours

Rapid-freeze protocol B

Freeze at −31°F (−35°C) or below until solid, then transfer to storage at −4°F (−20°C) or below for a minimum of 24 hours. A hybrid approach: rapid freezing at −35°C ensures complete solidification, after which −20°C storage for 24 hours achieves the required kill.

⚠

What Does NOT Work

Insufficient parasite controls

Standard home freezer (−18°C): does not meet the −20°C threshold. Brining or pickling: reduces but does not eliminate the hazard — Anisakis larvae survive 28 days in 80° salinometer brine. Marinating in citrus juice (ceviche): does not destroy Anisakis. Candling and physical inspection: reduces parasite numbers but does not eliminate the hazard. Evisceration (gutting): removes visceral larvae but not larvae already in the flesh — and if the fish was stored whole above 2°C before gutting, post-mortem migration into muscle has already begun. None of these alone is an acceptable substitute for compliant freezing.

The Formulated-Feed Aquaculture Exemption

The most important and most misunderstood provision of 21 CFR 123 Chapter 5 is the formulated-feed exemption. The FDA guidance states, verbatim:

"Species that normally have a parasite hazard as a result of consuming infected prey apparently do not have the same parasite hazard when raised only on pelleted feed in an aquaculture operation. You need not consider such aquacultured fish as having a parasite hazard."

— FDA Fish and Fishery Products Hazards and Controls Guidance, Chapter 5: Parasites (21 CFR 123)

The logic follows directly from the Anisakis lifecycle. The worm's larvae enter fish via infected marine prey — infected copepods, krill, and small fish eaten in the wild. A fish that eats nothing but heated, extruded pellets never encounters infected prey. The first step of the transmission chain is eliminated at the source. This is not a regulatory exception granted for convenience — it is a recognition that the biological hazard pathway structurally does not exist for these fish.

The exemption has important limits. The FDA is specific about when it does not apply:

A responsible purveyor verifies the specific culture methods of their aquaculture producers and does not apply the exemption without that verification.

Super-Freezing vs. Standard Freezing — Quality and Safety

FDA HACCP sets a floor for parasite destruction, not a ceiling. The three compliant protocols all kill parasites — but they differ significantly in their effect on fish quality.

Standard commercial freezing at −20°C (the Protocol ① minimum) freezes fish slowly enough that water migrates out of muscle cells before solidifying. The ice crystals that form between cells rupture cell membranes. On thawing, flavor compounds and soluble proteins leak out as drip loss — leaving a drier, softer, less flavorful fish. Fish oils (omega-3 fatty acids) remain liquid to approximately −110°C, so lipid oxidation continues even at −20°C storage, building rancid off-flavors over time.

Industrial super-freezing at −60°C to −86°C freezes so rapidly that ice crystals form inside cells, not between them — membranes remain largely intact, drip loss is minimized, and lipid oxidation is dramatically slowed. Super-frozen fish of peak quality is nearly indistinguishable from fresh when thawed correctly. Super-freezing exceeds all three FDA HACCP thresholds by a wide margin.

The practical implication: two pieces of tuna both labeled "sashimi-grade" may be very different products. One was frozen at −20°C (meeting the minimum, but cell membrane damage occurred); the other was super-frozen at −60°C (meeting the minimum and preserving structure). The label doesn't tell you which. At Sashimi DC, Bluefin Tuna and Sasshu Salmon are served fresh — freezing is not needed and never applied.

Parasite Status of Sashimi DC Products

Product-by-product parasite assessment

How each product is handled

Sasshu Salmon — Exempt

Raised in closed land-based tanks in Kagoshima fed exclusively on formulated feed. No ocean access. No wild prey. The Satsuma Sendai Unagi facility uses a constant-flow (かけ流し) mineral groundwater system; water is continuously replaced, eliminating ocean pathogen and parasite exposure. Qualifies under WAC § 246-215-03425(2)(d) / FDA Food Code 3-402.11(B)(2) formulated-feed exemption. Served fresh, never frozen.

Bluefin Tuna — Goto Islands

Farm-raised from wild-caught juvenile seed stock in ocean net pens off Goto Islands, Nagasaki. Feed is pre-frozen mackerel (from local fishers) stored in Hosei Suisan's freezer facilities — freezing destroys any parasites in the feed before it reaches the fish. Qualifies for the aquaculture exemption under WAC § 246-215-03425(2)(d) and FDA Food Code 3-402.11(B)(2): net pen + feed containing no live parasites = exempt. Documented per shipment via non-parasite declaration on the invoice/packing list. Served fresh, never frozen.

Hotaruika — Boiled at Source

Watasenia scintillans (firefly squid) can carry larvae of the nematode Crassicauda in the viscera — the adult form parasitizes beaked whales. Recent surveillance data (IASR 2025) places infection rates at 0.6–1.6%, down from 2–7% in pre-2011 studies. Raw visceral consumption carries risk of cutaneous creeping eruption and intestinal obstruction. Sashimi DC's hotaruika is lightly boiled at source in Toyama immediately after landing — the standard Japanese preparation — destroying nematode larvae while preserving texture.

Hokkaido Uni — No Parasite Risk

Sea urchin is an echinoderm, not a fish. The Anisakis lifecycle does not involve sea urchin as a host. FDA HACCP parasite-destruction rules under 21 CFR 123 are designed for fish and fishery products carrying nematode risk — they do not apply to uni. The primary food safety concern with uni is bacterial spoilage from temperature abuse, not parasites. Hokkaido Uni at Sashimi DC is kept at strict refrigeration from Japan to your door.

Sources

Full reference list with citations →