The 43 Per Cent of the IVD Market That Files No Accounts

Where the number comes from, and what is inside it

The global in vitro diagnostics market is USD 108.0 billion in 2025 in our construction. Of that, USD 61,525 million — 56.97% — is the disclosed, IVD-attributable revenue of twelve companies, taken from their own filings and results releases. The remaining USD 46,475 million — 43.03% — is a stated residual.

That residual is not a rounding bucket and it is not an allowance for error. It is a list: Chinese and regional reagent manufacturers, of which the Chinese domestic industry is the largest single block; Japanese, Korean, Indian, Brazilian and Turkish suppliers serving their own health systems at prices the multinationals do not reach; hospital in-house and consolidated laboratory networks; blood banking and immunohematology; unlisted specialists; and the IVD lines of named vendors that are not separately disclosed.

That last item deserves its own sentence, because it is the one readers least expect. Danaher, Thermo Fisher, Abbott and Siemens all report diagnostics inside a larger structure, and only part of what they sell is clinical diagnostic reagent. A bottom-up build of this market therefore cannot be a clean sum of audited diagnostics lines even for the largest companies in it. Some of the 43% is not small companies. It is big companies declining to break out a line.

The invisible block sets the visible price

Here is the part that changes how the residual should be read. The Chinese domestic reagent industry inside that 43% supplies its own health system at prices the multinationals do not reach, and centralised volume-based procurement turned that fact into an administered price. The Jiangxi biochemical alliance awarded 2,314 specifications at an average reduction of 70%. Tumour markers and thyroid function were tendered across 28 provinces in 2024, cardiac markers under Anhui’s leadership, routine blood, stool and urine reagents under Gansu in 2025, and Jiangsu brought a testing service inside the mechanism with its non-invasive prenatal testing award from 1 January 2025.

The consequence shows up in the visible 57%. Roche’s diagnostics sales in China fell 24% while the same division grew 7% at constant exchange rates excluding China. Siemens reported a low-double-digit decline in China diagnostics revenue and named centralised volume-based public procurement as the cause. Sysmex cut guidance from JPY 560 billion to JPY 535 billion citing Chinese government procurement policy and US tariffs. Werfen grew 6% organically excluding Asia against 3.3% overall.

So the residual is not the part of the market that does not matter. It is the part that sets the price at which the whole market transacts, and it is the part for which almost no audited figure exists. Any estimate of in vitro diagnostics is therefore an estimate of the least observable block, disciplined by the most observable one — and it should say so.

The residual grows faster, and stays large

Our base assumes the residual grows 3.1% in 2025 against 1.2% for the named pool, for a reason that is visible in the disclosures: the tail is where volume growth is fastest and where price deflation is most severe. Combined, the two pools produce 2.02% of market growth on a USD 108,000 million base.

Compound that residual at 3.1% to 2033 and it reaches USD 59,332 million. Against a 2033 market of USD 168,814 million that is 35.1% — more than a third of this market in 2033 still sitting behind no audited account, after eight years of forecast.

This is where the error bar in every IVD figure actually lives. A single percentage point on the residual’s growth assumption moves the base year by USD 464.8 million before it compounds at all, and by considerably more by 2033. Two estimates that differ by two or three per cent of market size can be produced honestly from the same filings, purely by disagreeing about a block nobody discloses.

The disagreement in this market is about scope, not method

Three boundary decisions do most of the work in explaining why published numbers for this market differ by tens of billions of dollars, and none of them is a methodological dispute.

Laboratory services in or out. Ours are out. A laboratory performing a test is selling a service; the reagent is an input to it. That single decision is why a figure of USD 108 billion and a figure of roughly USD 160 billion can both describe diagnostics honestly — one is the reagent and instrument market, the other is the testing market, and the laboratory that converts a reagent into a reportable result keeps the difference between them.

Research-use-only reagents in or out. Ours are out. RUO reagent sold to a life-sciences laboratory is not a diagnostic, even when it is the same chemistry in the same box, and including it inflates the market with revenue that carries none of the regulatory cost of a clinical assay.

The vendors’ non-clinical businesses in or out. Ours are out, and this one is materially understated in most summaries. Excluding industrial and pharmaceutical quality control removes roughly 20% of bioMerieux’s reported sales and a material slice of Thermo Fisher’s microbiology franchise. Those numbers are reported, they are large, and they are not diagnostics in the sense a laboratory buyer means.

A reader comparing two estimates for this market that differ by more than the forecast range should assume a boundary difference until proven otherwise. It is almost never the growth rate that disagrees at that scale; it is the perimeter.

What our interviews say about the tail

The tail is the hardest part of this market to observe and, in our primary layer, the part practitioners describe most consistently. Three points recurred across depth interviews with laboratory directors, distributors and procurement staff, a procurement-side survey, and conference-floor conversations in a dozen markets.

The tail is price-anchored, not cost-anchored. Local manufacturers quote against the multinational list price minus a discount rather than against a built-up cost. That is why an administered price can move the whole market’s level in a single tender round: the reference point is visible to every bidder, and it is the incumbent’s list.

Local manufacture is a procurement criterion in its own right. Domestic-preference rules, localisation requirements and tariff exposure change the relative cost of an imported reagent faster than any registration decision changes anything. Share in these markets has moved on a factory being built, not on an assay being cleared.

Nobody in the tail publishes revenue. The only routes to sizing it are procurement volumes and distributor structures, both of which are partial and both of which are models. That is not a reason to leave the block out of the market. It is a reason to state it as a residual with an assumed growth rate, which is what makes it checkable and falsifiable rather than invisible.

Three questions to ask of any IVD number

Instruments and reagents, or testing? Research-use-only in or out? And what residual, at what growth rate? A forecast that does not state a residual is not a forecast of the market — it is a forecast of the companies that file, presented as a forecast of the market. Ours states 43.03%, growing 3.1% in the base year, and that is deliberately the most conspicuous number in the base construction.

It also has a use beyond hygiene. Since the residual is where the price discipline originates, the growth rate assumed for it is a statement about competition: assume the tail grows slowly and you are assuming the administered price continues to bite; assume it grows quickly and you are assuming domestic volume outruns domestic price deflation. Readers who disagree with our 5.74% will usually find they disagree about the tail, not about the twelve companies whose accounts they can read. The full market report shows both pools separately for exactly this reason, and the diagnostic tests and virus testing kits reports carry the same perimeter discipline at discipline level.

Companion analyses in this cluster: China’s Price Cut Is Worth Half a Point of the IVD Forecast, Not Four puts a value on the base year; Two Diagnostics Businesses, the Same Operating Profit, a Quarter of the Revenue reads the margins of the four franchises that do disclose them. Both sit under the Global In Vitro Diagnostics (IVD) Market report, 2026-2033, which keeps the named pool and the residual separate throughout, and alongside the home-based molecular testing and virus testing kits reports, where the same perimeter rules apply at discipline level.

About the method

Roughly 70% secondary and 30% primary. The base year is built bottom-up from disclosed IVD-attributable revenue — twelve participants, USD 61,525 million, 56.97% — plus a stated residual of USD 46,475 million, 43.03%; no published aggregate was used to set the level. Non-USD revenue is translated at disclosed 2025 average rates. The primary layer is used to explain mechanism — how a tender is sequenced, why local manufacture acts as a procurement criterion, why the tail prices against the incumbent list — and never to carry a figure.

Construction choices, the disclosure table, the segment build and the sensitivity cases sit in the Global In Vitro Diagnostics (IVD) Market report, 2026-2033. Two companion analyses follow from it: China’s Price Cut Is Worth Half a Point of the IVD Forecast, Not Four, on what the base year is worth in a compound rate, and Two Diagnostics Businesses, the Same Operating Profit, a Quarter of the Revenue, on the part of the market that does publish a margin.