This is the playbook I’d use if you called me on a Tuesday morning with a partial fail and a full production calendar behind it. It’s practical, it’s regulator-friendly, and it keeps your options open.
Why partial lot microbial failure shows up even when most of the harvest is “fine”
Here’s the part that trips people up: most state programs are not testing every single unit. Your lab is pulling a representative sample, and in many cases that sample is composited across the lot. If your lot isn’t truly uniform, the result won’t be either.
Microbial load is rarely evenly distributed. One rack dries slower. One tote sits in a warmer corner. One trim shift keeps lids open longer during staging. Small differences stack up. If you want a deeper look at how “mixed-risk” lots happen in real post-harvest workflows, I laid it out in our guide on mixed-risk cannabis lot handling.
Step 1: Quarantine fast (yes, treat partial lot microbial failure like the whole lot is implicated)
When a fail hits, your best move is boring and immediate: put the full lot on hold. No transfers. No pre-roll conversion. No extraction runs. No “just this one thing” exceptions. Even if you’re sure the issue is limited to one tote, regulators expect you to preserve the situation as it existed when the sample was collected.
Quarantine is less about panic and more about preventing new questions. What moved? What got blended? What got relabeled? The cleaner your first response, the easier your retest or remediation pathway will be. If you need a simple framework for building this into your routine, you’ll like our post on a batch release plan built around microbial risk.
- Physically isolate the inventory in a locked cage, controlled room, or clearly marked zone.
- Freeze work orders tied to that lot in your seed-to-sale system.
- Document the scene right away: label photos, storage layout, environmental readings, and access logs.
Step 2: Investigate before you pick a cannabis retesting strategy
I know the impulse: “Let’s just retest.” But if you retest before you understand where the variability came from, you can burn your one best shot at a clean, defensible outcome. A partial failure is usually pointing to a localized condition or a specific handling moment.
Start with a tight problem statement. What failed, and by how much? Which product form? When was the sample taken? Then walk backward through your timeline: harvest, bucking, drying, curing, trimming, storage, staging. This is where you find the weak link.
If you want a structured CAPA-style approach, I use this same flow in our root-cause post on microbial root cause analysis after a failure.
- Map the lot on paper first: room, rack, cultivar, harvest date, tote ID, trim shift, storage location.
- Pull moisture and water activity data (or measure now if you can). Moisture history is the usual suspect.
- Look for risk moments: open-lid staging, warm holds, slow dry-down zones, condensation events, dense packing.
Even if your state doesn’t allow partial release, your documentation still matters. It supports your next compliance step, and it gives you a real fix instead of a temporary pass.
Step 3: Partial lot microbial failure does not mean you can freely split inventory, so check cannabis lot segregation rules
Let’s say you’ve identified the likely “hot spot” and you’re confident most of the lot is clean. You still can’t assume you can ship the good portion. Cannabis lot segregation is tightly regulated, and the rules vary a lot by state and by product category.
Some programs treat the lot as the compliance unit until you follow a specific process for re-sampling, reanalysis, remediation, or reprocessing. Washington is one example where the rule language is pretty direct about how remediation and retesting are handled under WAC 314-55-102, which you can read at Washington’s administrative code site.
So do this in order:
- Confirm which pathways your state allows for the specific failure type.
- Confirm whether you can legally create a new lot, and what documentation is required.
- Confirm who must collect samples and how chain of custody works.
If you skip this and “get creative” in the warehouse, you can turn one microbial issue into a traceability issue, and those are harder to recover from.
Step 4: Build a defensible cannabis retesting strategy (don’t waste your one clean retest)
Your cannabis retesting strategy should match what your regulator actually recognizes. Many states distinguish between reanalysis and retest. Reanalysis is the lab rerunning the same sample. Retest is pulling a new sample. Those are not the same thing, and they don’t carry the same weight.
Oregon is a good example of a state that has laid out a more structured set of options, and SC Labs has a useful plain-language breakdown you can reference at SC Labs’ summary of Oregon microbial retest options.
- Confirm eligibility before you do anything else. Some failure types have limited paths.
- Protect chain of custody. Don’t self-sample unless your state explicitly allows it and your SOP supports it.
- Sample with intention. If you suspect the issue is localized, a “grab sample from the easiest tote” is a fast way to fool yourself.
One tip I give clients all the time: if your internal notes show one area tends to run wetter, treat that as your worst-case area. You want a passing result because you controlled the conditions, not because you accidentally sampled the safest corner. I know that sounds picky, but it’s how you avoid the dreaded re-fail.
Step 5: Decide whether remediation makes sense for partial lot microbial failure (and for what actually failed)
If remediation is allowed in your state for the failure type and product form, you still have a business decision to make. It’s not only “can you save it,” it’s whether you can save it profitably and still feel good about the finished quality.
When you’re deciding, a simple cost matrix helps. Canna Labs Consulting has a straightforward framework you can use to compare remediation versus destruction at their remediation decision matrix article.
- Remediation cost: labor, downtime, packaging, logistics, equipment time or service fees.
- Retesting cost: required scope plus sampling fees.
- Quality impact: weight loss, terpene loss, product positioning changes.
- Time impact: delayed revenue, missed windows, cashflow pressure.
If the failure is pathogen-related, like Aspergillus in certain markets, slow down and align with your state’s expectations. A kill step can reduce viable organisms, but your program may require specific retesting scope and documentation. This is one of those areas where being “sorta right” isn’t good enuf.
Step 6: Prevent the next partial lot microbial failure with a few tight, repeatable controls
Once the immediate hold is resolved, you’ve got a chance to make the next run smoother. Partial failures usually point to uneven dry down, inconsistent moisture gating, or sampling habits that don’t reflect how the lot is really distributed.
I also recommend you don’t close out your CAPA just because one retest passed. A meaningful fix holds across multiple lots, and you can see the trend improve. If you want a practical CAPA structure built for microbial stability, it’s outlined here: our CAPA program framework for microbial failures.
- Set internal action limits below the state threshold so you catch drift early.
- Standardize composite sampling across multiple sub-locations, not just the most convenient spot.
- Add moisture and water activity gates before long storage or packaging.
- Use a validated kill step when your risk profile and economics justify it.
If you want to reduce bioburden proactively, our ozone-based WillowPure systems are designed to function as a scalable kill step for flower and trim. You can see the system options at WillowPure systems. If what you really need is help designing the SOPs, sampling plan, and CAPA workflow around your facility, our team does that too at Willow consulting services.
FAQ: Partial lot microbial failure
Can you release the “good” portion if only part of the lot failed?
Usually not unless your state has a specific, state-approved pathway for doing it. In many programs, the compliance status applies to the entire lot as defined in your tracking system. You typically need to follow formal rules for retesting, remediation, reprocessing, or legally creating a new lot.
Should you retest immediately after a partial lot microbial failure?
Not before you quarantine and document, then confirm eligibility under your state rules. If you rush into a retest without aligning sampling and chain-of-custody requirements, you can waste money and still end up with a result that won’t hold up.
What should you investigate first during a partial failure?
Start with moisture history and handling variation. Check differences by room, rack, tote, cure conditions, and staging steps that may have created warm or humid pockets. Then compare that to what was sampled and when.
Is cannabis lot segregation worth doing operationally?
It can help you isolate risk and prevent cross-contamination, but only if it’s legal in your state and fully traceable. Segregation that isn’t defensible on paper can create bigger compliance problems than the original microbial fail.
What is the simplest way to reduce future microbial failures?
Put earlier warning systems in place: internal action limits, consistent composite sampling, and clear moisture or water activity gates. If your volume and risk profile justify it, a validated kill step can turn microbial compliance into a controlled process instead of a scramble. If you’re stuck, ask for help sooner than later. It’s almost always cheaper.
Conclusion: protect options first, then fix the system
When partial lot microbial failure lands on your desk, discipline beats speed. Quarantine the full lot, document what you have, investigate where variability entered the process, then choose a compliant pathway for cannabis lot segregation, remediation, or your cannabis retesting strategy. If you want a second set of eyes on your sampling plan or your post-harvest controls, reach out to me at Willow. You’ll get straight answers and a workflow you can actually run day after day, not a binder that collects dust.