Salmonella Controls That Prevent Chocolate Recalls

A recent FDA recall expansion illustrates exactly the kind of failure that a well-structured food safety system is designed to catch before product ever reaches consumers.

In May 2026, Spring & Mulberry expanded its voluntary recall of select chocolate bars due to possible Salmonella contamination — an expansion of a previously announced action, which means the initial recall did not fully contain the problem. The FDA notice, published May 8, 2026, covers product distributed out of Raleigh, NC. You can review the agency's original posting at FDA's recall database.

The recall itself is not the story I want to tell here. The story is what a manufacturer would have needed to have in place — and actually running — to prevent this situation entirely. Because Salmonella in chocolate is not a new problem, and the controls that stop it are well understood. They just have to be implemented seriously.

Why Chocolate Is a Surprisingly High-Risk Environment for Salmonella

Most people associate Salmonella with raw poultry or undercooked eggs. Chocolate feels like the opposite of a high-risk food — it's low moisture, shelf-stable, typically not considered a ready-to-eat hazard in the same breath as deli meat. That intuition is wrong, and it has caused real harm.

Chocolate's low water activity (generally below 0.60) does not kill Salmonella. It preserves it. Studies have shown Salmonella can survive in chocolate for months or even years at ambient temperature. The low-moisture environment actually protects the pathogen from the thermal interventions that might otherwise work. And because chocolate is often consumed directly without any kill step by the consumer, contamination at any point in the manufacturing process — from incoming cocoa or dairy ingredients to post-process handling — carries real risk.

According to CDC data, Salmonella causes approximately 1.35 million infections, 26,500 hospitalizations, and 420 deaths in the United States annually, making it one of the most consequential foodborne pathogens the food industry manages. For immunocompromised individuals, the elderly, pregnant women, and young children, even a low infectious dose can be serious.

The FDA has documented multiple Salmonella outbreaks traced to chocolate and cocoa products over the past two decades, including a significant 2006 outbreak linked to contaminated cocoa powder that affected consumers in multiple countries. The hazard is real, documented, and preventable.

The Regulatory Framework: What FSMA Actually Requires

The Food Safety Modernization Act (FSMA) and its implementing regulations are the primary legal architecture governing this space. For chocolate manufacturers, the most directly applicable rule is 21 CFR Part 117 — Current Good Manufacturing Practice, Hazard Analysis, and Risk-Based Preventive Controls for Human Food (the Preventive Controls for Human Food rule, or PCHF rule).

Under 21 CFR § 117.130, manufacturers are required to conduct a hazard analysis that identifies known or reasonably foreseeable hazards — including biological hazards like Salmonella — and determine which hazards require a preventive control. Salmonella in a chocolate manufacturing environment almost certainly requires a preventive control under this framework. That is not a gray area.

21 CFR § 117.135 then requires that those preventive controls be implemented and managed. For a pathogen like Salmonella in a low-moisture food, the relevant preventive controls typically include:

• Process controls: validated kill steps (usually a roasting or conching process applied to cocoa)
• Sanitation controls: documented environmental monitoring programs designed to detect and eliminate harborage points
• Supply-chain controls: verified supplier programs under 21 CFR § 117.410–117.475 for ingredients that could introduce the hazard

And 21 CFR § 117.150 requires monitoring — actual documented evidence that the controls are working, on a defined schedule.

When a recall expands, as this one did, it is often a signal that at least one of these elements was either absent, not fully implemented, or not functioning as designed. An initial recall that has to be expanded suggests the scope of the contamination was not fully understood at the outset — which points back to gaps in traceability, environmental monitoring, or both.

What an Effective Hazard Analysis Would Have Identified

A proper HARPC (Hazard Analysis and Risk-Based Preventive Controls) document for a chocolate manufacturer should identify Salmonella as a hazard requiring a preventive control at several points in the process flow:

Incoming cocoa and dairy ingredients. Cocoa beans are an agricultural commodity with documented Salmonella prevalence. Raw cocoa and cocoa-derived ingredients should trigger supplier verification activities under 21 CFR § 117.410. This means onsite audits or third-party audit results, ingredient testing (either by the supplier with verification records, or by the manufacturer), or a combination. A certificate of conformance alone is generally not sufficient verification for a pathogen hazard in a high-risk ingredient.

In-process kill step. Roasting is the primary kill step for Salmonella in cocoa. The roasting process should be validated to demonstrate it consistently achieves a sufficient log reduction — typically a 5-log reduction is the accepted standard for Salmonella in low-moisture foods. 21 CFR § 117.160 requires that process preventive controls be validated, meaning there should be scientific or technical evidence on file that the roasting parameters (time, temperature, airflow) actually achieve the intended pathogen reduction. If that validation study was never done, or was done on a different product formulation or line configuration, the control is not properly validated.

Post-lethality environment. This is where chocolate manufacturing gets complicated. After the kill step, product moves through cooling, tempering, enrobing, and packaging — all of which involve exposure to the manufacturing environment. Any harborage point for Salmonella in that post-lethality zone is a recontamination risk. Environmental monitoring, using programs sometimes called Environmental Monitoring Programs (EMPs), are the tool designed to detect harborage points before they contaminate product.

A well-run EMP collects swabs from equipment surfaces, floors, drains, walls, and air handling components on a rotating schedule — and importantly, maps and investigates findings rather than just logging them. Finding Salmonella in the environment is not automatically a failure. Failing to investigate and correct it is.

The Environmental Monitoring Program Gap

In my experience reviewing food safety programs, the EMP is one of the most commonly underimplemented preventive controls. Manufacturers set one up to satisfy an audit, collect swabs on schedule, and then treat the data as a filing exercise rather than an early warning system. That is not an EMP — that is a compliance theater exercise.

An effective environmental monitoring program for a chocolate facility should include:

The key thing about trend analysis is that it turns individual data points into a picture. A single positive in Zone 3 may not mean much on its own. Three positives in Zone 3 near the same drain over six months is a pattern — and that pattern tells you where to look before product is affected.

FDA investigators reviewing environmental monitoring records during an inspection will look at all of these elements. If the records show testing but no trend analysis, no corrective action documentation, and no evidence of program expansion after positive finds, that is a significant gap — and it is the kind of gap that tends to surface in 483 observations and warning letters.

Supplier Controls: The Incoming Ingredient Problem

One of the things a recall expansion often signals is that the source of contamination was not fully traced during the initial action. That tracing problem is usually a symptom of inadequate supply chain controls or lot traceability records.

21 CFR § 117.410 requires manufacturers to establish and implement a supply-chain program for ingredients where the hazard requires a supply-chain-applied control. For cocoa and cocoa-derived ingredients, the pathogen hazard is known and documented, so supplier controls are not optional.

What that actually looks like in practice:

• Approved supplier list with documented qualification criteria (audit results, testing history, certifications)
• Supplier questionnaires and/or onsite audits on a defined frequency — typically annual for high-risk ingredients
• Certificate of Analysis review for each incoming lot, with defined acceptance criteria
• Periodic incoming material testing for Salmonella, particularly for high-risk ingredients like cocoa powder or milk powder
• Traceability records that link each ingredient lot to specific finished product lots

That last point — traceability — is what makes a recall manageable when it does happen. Without solid lot-to-ingredient linkage, a manufacturer cannot confidently define the scope of affected product, which is exactly the situation where an initial recall gets expanded.

The FDA's traceability rule (21 CFR Part 204, implementing FSMA Section 204) has additional specific requirements for certain foods. While the compliance dates for some firms have been extended and FDA issued a final rule in 2022 with subsequent guidance, manufacturers of products with a Salmonella risk would be well advised to have robust traceability systems in place regardless of whether their specific product is on the Food Traceability List.

What a Recall Expansion Actually Tells You

When a voluntary recall gets expanded — as happened here — a few explanations are possible, and most of them point to systemic gaps rather than isolated incidents:

Scenario 1: The scope of affected lots was underestimated initially. This usually means lot-level traceability records were incomplete or the distribution records were not granular enough to identify all affected product at the time of the first action.

Scenario 2: New information surfaced about the source of contamination. The manufacturer found contaminated product or environmental positives that extended the affected timeframe or product range beyond the initial recall scope.

Scenario 3: The root cause investigation revealed that the problem was more widespread than the initial sampling suggested. This is common when environmental Salmonella is the source — the organism can be present in multiple zones before it's detected.

In any of these scenarios, the underlying prevention question is the same: what would have caught this earlier? A robust EMP with trend analysis catches environmental harborage. Lot-level traceability catches the scope problem. Validated supplier controls reduce the probability that contaminated ingredients enter the facility in the first place.

Practical Steps Food Manufacturers Should Take Now

This recall is a timely reminder to review your own systems. If you manufacture chocolate, low-moisture foods, or any product where Salmonella is a known biological hazard, here is where I would start:

1. Review your hazard analysis. Is Salmonella identified as a hazard requiring a preventive control? Is the basis for that determination documented? If your hazard analysis was completed before your current product line or facility configuration, it needs a reassessment.

2. Validate your kill step. Do you have a validation study on file that demonstrates your roasting or other thermal process achieves the intended Salmonella reduction? Is that validation product-specific and equipment-specific? Validation is required under 21 CFR § 117.160 and should not be delegated to "we follow industry standard practice."

3. Audit your EMP. When did you last conduct a trend analysis of your environmental monitoring data? Do your corrective action records show genuine root cause investigation, or just re-cleaning and re-testing? Are Zone 1 findings triggering hold-and-test protocols?

4. Verify your supplier controls. For your highest-risk ingredients, do you have recent audit results or testing data on file? When was the last time you verified that your suppliers are actually implementing the controls you're relying on?

5. Test your traceability. Run a mock recall. Start with a specific finished product lot and trace it back to ingredient lots and forward to distribution records. See how long it takes and whether the data is complete. The FDA expects manufacturers to be able to execute a recall traceability exercise in two hours or less.

A well-designed food safety program does not guarantee that contamination never occurs — raw agricultural ingredients carry biological hazards that cannot always be fully controlled at the source. What a well-designed program does is catch the problem early, contain it to a defined scope, and prevent the kind of recall expansion that turns a manageable situation into a significant one.

The Broader Pattern in FDA Enforcement

Salmonella in low-moisture foods has been a sustained area of FDA focus. The agency issued a compliance policy guide on environmental monitoring (CPG Sec. 555.320) and has published its own Environmental Monitoring for Microbial Pathogens: Food and Environmental Pathogen Laboratory Methods guidance. Following a 2009 peanut paste outbreak that led to one of the largest food recalls in U.S. history (affecting over 3,900 products), FDA significantly elevated its expectations for environmental monitoring in facilities handling low-moisture foods.

Manufacturers who have not updated their food safety programs to reflect the current state of FDA expectations — including the Preventive Controls rule requirements, the Supply-Chain Program requirements, and the Environmental Monitoring guidance — are operating with compliance risk that will eventually surface, either in an inspection or in a product safety event.

In my view, most recalls that expand are not the result of bad intent. They are the result of systems that were built to pass audits rather than to actually work. There is a meaningful difference between the two, and that difference tends to show up at exactly the wrong time.

If you are reviewing your food safety program or preparing for an FDA inspection, Certify Consulting offers full-service compliance consulting with a 100% first-time audit pass rate across 200+ clients. Learn more about our food safety and FDA compliance services.

Last updated: 2026-06-02