5 Stability Tests Korean ODM Labs Run Before a Formula Ships
A look at the accelerated-aging protocol that decides whether a K-beauty formulation is ready for production — and the five gates a reputable Korean ODM partner will document for your batch.
Key Takeaways
Cosmetic stability testing in Korean ODM labs typically uses an accelerated-aging chamber at 40 to 45°C for 90 days. The Korean MFDS reference protocol for functional cosmetics is 40°C / 75% relative humidity for 90 days.
Five tests recur across most Korean ODM stability protocols: centrifuge for emulsion stability, pH drift, viscosity drift, microbial enumeration, and sensory consistency.
Centrifuge testing is commonly cited at 3,000 rpm for 30 minutes — a standard physical-stability screen for emulsions.
pH drift is generally accepted within ±0.5 units across the test period; viscosity tolerance is generally accepted within ±15 percent.
The microbial limit for general leave-on cosmetics under ISO 17516:2014 is ≤10³ CFU/g, with a stricter ≤10² CFU/g for products used around the eyes, on children under three, or on mucous membranes.
Stability protocols vary by lab. The presence of a documented protocol — not the marketing claim — is the diligence signal.
Indie brand founders should request the stability test certificate as part of the sample handoff. Reputable Korean ODM labs treat this as a routine deliverable.
Why accelerated stability testing exists
Skincare formulations need to hold their physical, chemical, and microbiological properties across the time between manufacture and the consumer's last use. For products with a 24-month claimed shelf life, real-time stability testing — sitting on a 25°C shelf for two years — is the gold-standard confirmation. But brands cannot wait two years to launch.
Accelerated stability testing is the workaround. The principle is the Arrhenius equation: chemical reaction rates roughly double for every 10°C increase in temperature. A formulation pushed to 40 to 45°C for 90 days has been stressed harder than the same formulation sitting at room temperature for roughly 12 to 18 months. If the formulation is going to break — emulsion separation, pH drift, preservative failure, color or odor change — it usually breaks inside the chamber.
Korean MFDS specifies 40°C with 75 percent relative humidity over 90 days as the accelerated condition for new sunscreens and many functional cosmetic categories. ISO/TR 18811:2018 and the related cosmetic stability literature describe similar conditions in the 40 to 45°C range. Many Korean ODM labs run their internal screen at 45°C as a stricter condition than the regulatory minimum, then maintain a parallel real-time study at 25°C for the full claimed shelf life.
Five tests recur in the Korean ODM accelerated stability panel. None of them are unique to Korea — they trace back to ISO and pharmacopeial standards used worldwide — but the discipline with which a quality-focused Korean ODM applies them is what separates a serious manufacturing partner from a contract filler.Test 1: Centrifuge — preliminary emulsion screen
The protocol: Most cosmetic emulsion stability literature describes a screen at 3,000 rpm for 30 minutes, often performed at slightly elevated temperature to accelerate creaming or coalescence behavior. Some labs run additional speed tiers (1,000, 2,000, 3,000 rpm at shorter durations) to bracket the failure point.
What it measures: Whether the emulsion holds. Most skincare products combine a water phase with oil-soluble actives, emollients, or active ingredient carriers. Centrifugation applies hundreds to roughly a thousand times the force of gravity, compressing months of natural settling into roughly half an hour.
Pass condition: No visible phase separation. No oily creaming layer at the top. No clear watery layer at the bottom. No flocculation or sedimentation of the dispersed phase.
A failed centrifuge result is one of the most common reasons a Korean ODM rejects a batch on day one. The fix is usually emulsifier system rebalancing — adjusting the HLB value, switching to a co-emulsifier system, or adding a polymer stabilizer like xanthan gum or carbomer to slow phase separation kinetics. Korean labs that specialize in lightweight, "watery" K-beauty textures lean heavily on cold-process emulsion technology, which makes the centrifuge gate more demanding rather than less.
The centrifuge test is a preliminary screen, not a complete stability assessment. It addresses physical emulsion stability only. A formula that passes centrifuge can still fail on pH, viscosity, microbial count, or sensory checks across the full 90-day chamber test.
Test 2: pH stability across 90 days
The protocol: pH is measured on day 0 and at predefined intervals (commonly day 15, 30, 60, and 90) using a calibrated probe at the same temperature each time.
What it measures: The chemical stability of the formulation. pH drift is a leading indicator of ingredient breakdown. If a vitamin C serum's pH rises across the test period, the ascorbic acid is oxidizing. If a niacinamide toner's pH rises, the niacinamide may be partially converting to nicotinic acid — a transformation that triggers facial flushing in some users. If a cream's pH falls, the carbomer or other acidic-stable polymers may be hydrolyzing.
Pass condition: pH change generally within ±0.5 units across the test period. Some manufacturers tighten the criterion (for example, ±0.3 or even ±0.1 units) when the formula targets a narrow window — sunscreens, retinoid serums, and acid exfoliants are common cases where the tighter gate is appropriate.
The pH window is not arbitrary. The skin's acid mantle sits between roughly pH 4.5 and 5.5. A formulation that drifts outside its intended window — into territory the skin reads as too acidic or too alkaline — does not just become less effective. It can become destabilizing to the skin barrier function the formulation was supposed to support.
Korean ODM labs control pH drift through chelating agents (disodium EDTA captures trace metal ions that catalyze oxidation), antioxidant systems (tocopherol; sodium metabisulfite for vitamin C formulas), and buffer systems (citrate, phosphate, or lactate buffers calibrated to the target pH). The ±0.5-unit gate is what separates a formulation that holds its window from one that quietly degrades on the bathroom shelf.
Test 3: Viscosity stability
The protocol: Viscosity is measured on day 0 and at the chamber checkpoints (commonly day 30, 60, and 90) using a rotational viscometer (Brookfield-style is the standard) at a fixed spindle speed and temperature.
What it measures: Whether the formulation feels the same on day 90 as it did on day 0. A toner that thickens into a gel during accelerated aging will land in the consumer's hand differently than the lab sample they signed off on. A cream that thins into a runny lotion has lost the carrier system that holds the actives in the upper layers of the skin.
Pass condition: Viscosity change generally within ±15 percent of the day-zero reading across all checkpoints. As with pH, the criterion can be tightened where the rheology is part of the brand promise — gel-cream textures, "splash" toners, and sunscreen fluids are categories where a smaller drift window is often specified.
Viscosity drift is often invisible to the formulator until the chamber surfaces it. Polymer thickeners (carbomer, xanthan gum, sclerotium gum) can swell or deflate as they hydrate over weeks. Some cellulose-based thickeners are sensitive to enzymes that survive insufficient sterilization. Gum-based texture systems can shift with temperature exposure during transit, even if the chamber test passed.
For a brand founder, viscosity drift shows up in customer reviews in one specific way: "It felt different from the sample." That is not a marketing problem. It is a stability problem.Test 4: Microbial count — to ISO 17516 limits
The protocol: Microbial enumeration following ISO 21149 (aerobic mesophilic bacteria) and ISO 18416 (Candida albicans), with screening for specified pathogens. Samples are plated on growth media and incubated under controlled conditions; colonies are counted at the endpoint.
What it measures: Whether the preservative system is doing its job. Skincare products are not sterile, but they cannot serve as breeding media. Every time a consumer's finger enters a jar, every time a pump introduces ambient air, the preservative system is asked to neutralize new microbial load.
Pass condition (per ISO 17516:2014):
Total aerobic mesophilic count ≤10³ CFU/g (1,000 CFU/g) for general leave-on cosmetics.
≤10² CFU/g (100 CFU/g) for products used around the eyes, on children under three, or on mucous membranes.
Specified pathogens — Staphylococcus aureus, Pseudomonas aeruginosa, Candida albicans, Escherichia coli — must be undetectable in the prescribed sample size.
These limits are international references. ISO 17516:2014 is the headline standard; US Pharmacopeia chapter 51 and EU Cosmetics Regulation 1223/2009 Annex I describe broadly compatible frameworks. Korean ODM labs that ship internationally typically design preservative systems to clear the strictest applicable limit by default, since one batch is likely to land in multiple regulatory regions.
Korean preservative system design has shifted significantly in the last decade. The older "parabens plus phenoxyethanol" baseline is increasingly replaced by "clean" preservation: ethylhexylglycerin, caprylyl glycol, sodium benzoate, potassium sorbate, and natural-derived alternatives. The trade-off is that clean preservative systems often have less kill margin than the older systems and require tighter pH control, packaging discipline, and stability testing to clear the same gate.
This is also the test most often associated with clean K-beauty positioning. A brand that wants to claim "no parabens" needs an ODM partner whose preservative system can clear ISO 17516 limits without parabens. That capability is not universal across factories.
Test 5: Sensory consistency — color, odor, application feel
The protocol: Visual color comparison against a sealed reference sample stored at low temperature (the cold reference is treated as the unaged baseline). Odor evaluation by trained evaluators against the same reference. Application-on-skin check at each checkpoint.
What it measures: Everything the instruments do not directly capture. A pH probe will not tell you that a cream has developed a faint metallic note at day 60. A viscometer will not flag that the color has shifted half a shade toward brown — a common signal of oxidation in formulations that use plant extracts. A centrifuge will not catch the slightly sticky aftertouch of a sunscreen whose silicone elastomer system has begun to phase out of the matrix.
Pass condition: Color falls within the visible-range tolerance against the cold reference, with no obvious shift noted by independent evaluators. No off-notes in odor at any checkpoint. Application feel matches the day-zero sample. Some labs apply quantitative color tolerances (a ΔE measurement against a reference) where the brand brief specifies a precise color target.
The sensory test is where formulation problems often surface first. Plant-derived ingredients (centella asiatica, green tea, propolis) are particularly prone to color drift as polyphenols oxidize. Essential-oil fragrance systems can develop "rancid" notes when the carrier oil oxidizes. Ferment ingredients (galactomyces, bifida) can develop a slightly sour edge at extended timepoints.
Korean labs typically maintain a sensory panel of two to four trained evaluators who run the day-30, day-60, and day-90 checks. Their notes go into the batch record alongside the instrument readings. A sensory failure does not always mean the formula is unsafe — it means it has drifted outside the brand promise. For a consumer, "the cream smells different from the one I bought last time" is a brand-equity loss that is difficult to recover.
What this means if you are evaluating a Korean ODM partner
Stability testing is invisible to most consumers and to most first-time brand founders. It happens inside the ODM lab, after the formulation handoff, before the first production run. By the time a brand is reviewing samples, the stability work has already been done — or it has been compressed, depending on the lab.
Three things are worth knowing:
One: Ask for the stability test report as a deliverable. A Korean ODM that takes the protocol seriously will produce a stability test certificate alongside the sample, summarizing the centrifuge result, pH and viscosity profiles across timepoints, microbial enumeration data with the applicable ISO 17516 limit, and sensory evaluation notes. If the report does not exist, or if the lab is reluctant to share it, that is a meaningful diligence signal.
Two: Accelerated testing predicts; real-time testing confirms. The 90-day chamber result is what releases the production batch on schedule. It is not a substitute for real-time stability data. For a product launching with a 24-month claimed shelf life, reputable Korean ODM labs typically begin a parallel real-time study at 25°C at the same time as the accelerated study. Asking whether the lab maintains a real-time arm is a useful diligence question.
Three: Five tests is a reasonable floor, not a ceiling. Premium Korean ODM labs add freeze-thaw cycling (alternating low and high temperatures to simulate transit through cold supply chains), light-exposure testing (UV cabinet for photo-sensitive ingredients like vitamin C, retinol, or some natural pigments), and packaging compatibility testing (the formula sitting in its actual primary packaging, not a generic glass beaker). For founders launching premium-positioned products, asking which of these supplementary tests are included in the standard protocol is worth the diligence time.
Stability protocols are not standardized across every Korean ODM. Some run the full panel as a routine deliverable. Others charge for stability testing as an add-on or limit it to abbreviated checks under time pressure. The protocol document itself — and the lab's willingness to share it — is the diligence signal. Marketing language about "rigorous testing" is not.
What a stability-tested formula looks like in your hand
The result of a clean five-test pass is a product that behaves consistently across its shelf life. The pump dispenses the same way at month one and month twelve. The texture lands on the skin the same. The color in the bottle has not deepened. The fragrance opens with the same top note. The active is still active.
This is the floor of what a serious Korean ODM partnership delivers. It is not a marketing claim. It is the basic obligation of the manufacturing process, and it is one of the reasons the Korean ODM channel — when run by a lab that takes the protocol seriously — has built its reputation in the indie K-beauty category.
For brand founders evaluating where to manufacture, the stability test report is one of the most boring documents in the entire process. It is also one of the most informative. Read it. Ask questions about it. The answer will tell you which lab is going to ship you a product that holds — and which is going to ship you a problem.
Frequently Asked Questions
How long does Korean ODM stability testing actually take?
Accelerated stability testing typically runs for 90 days at an elevated temperature (40 to 45°C, depending on lab and product type). Korean MFDS specifies 40°C with 75 percent relative humidity over 90 days for new sunscreens and many functional cosmetics. Real-time stability testing runs in parallel at 25°C for the full claimed shelf life of the product (commonly 24 to 36 months). The accelerated result drives the production schedule; the real-time result is the regulatory documentation.
Is the accelerated aging chamber the same everywhere?
The principle is broadly standardized — most cosmetic stability programs use 40 to 45°C as the accelerated condition. The execution varies. Reputable Korean ODM labs use temperature- and humidity-controlled cabinets with continuous logging, typically meeting ICH Q1A guidelines used in pharmaceutical stability work. Less rigorous labs may use simpler heated cabinets without comparable logging, which can compromise the validity of the test record.
What happens if a formulation fails one of the tests?
The batch goes back to the bench. The formulator identifies the failure mode (emulsion instability, pH drift, preservative inadequacy, sensory drift, etc.) and adjusts the formula. The revised formulation enters a new chamber test from day zero. This is the loop that delays product launches by a quarter or more, which is why experienced ODM labs front-load stability prediction during the formulation development phase.
Can I see the stability test report for a sample I receive?
Yes — and you should ask for it. A Korean ODM partner with proper documentation practice will produce a stability test certificate alongside the sample, summarizing the centrifuge result, pH and viscosity profiles across timepoints, microbial enumeration data with the applicable ISO 17516 limits, and sensory evaluation notes. If your ODM cannot produce this document, that is a diligence flag.Do natural and clean formulations require different stability protocols?
The protocol panel is similar; the challenge is harder. Natural-derived preservative systems often have less microbial kill margin than synthetic systems, so meeting ISO 17516 limits requires more careful pH control and packaging selection. Plant-extract pigments oxidize more readily, putting pressure on the sensory and color tests.
How does packaging interact with stability testing?
Packaging is part of the test. A formulation that passes stability in a glass beaker may fail when packaged in a plastic tube whose plasticizers leach into the formula at elevated temperature. Reputable Korean ODM labs running the full protocol conduct packaging-compatibility testing as a parallel arm.
Is stability testing different for sunscreen formulations?
Yes, in two ways. First, Korean MFDS specifies the 40°C / 75% RH / 90-day accelerated protocol for new sunscreens. Second, sunscreens add UV-exposure stability testing on top of the standard panel.
Build with a Korean ODM partner whose protocol you can read
The stability protocol is the work that separates a formulation that holds its promise across a customer's full bottle from one that quietly fails halfway through.
ALTA MEET connects U.S. and English-speaking brand founders with Korean ODM laboratories that maintain documented stability testing protocols, batch-level records, and the formulation depth to clear ISO 17516 limits and the typical pH and viscosity acceptance criteria on the first pass. Start with a free consultation to scope your product brief and review which labs match your shelf-life and clean-preservation requirements.
References: ISO 17516:2014 (Cosmetics — Microbiology — Microbiological limits); ISO 21149 (Enumeration of aerobic mesophilic bacteria); ISO 18416 (Detection of Candida albicans); ISO/TR 18811:2018 (Guidelines on cosmetic stability testing); ICH Q1A (R2) Stability Testing Guidelines; US Pharmacopeia chapter 51; EU Cosmetics Regulation 1223/2009 Annex I; Korea MFDS Cosmetic Safety Standards.
