Why Theranos Failed: The Blood-Test Startup That Skipped Proof

Theranos failed because its technology never worked as claimed, and it kept selling tests to patients and shares to investors anyway. The company said it could run hundreds of lab tests from a few drops of fingertip blood, but its own device ran only around a dozen tests at most, and the rest ran on conventional machines with diluted samples. When the Wall Street Journal exposed the gap in October 2015, regulators, partners and investors followed, and the company dissolved in 2018.

The scale of the miss is hard to overstate. Theranos raised roughly $900 million or more (reported) and reached a valuation of about $9 billion in 2014. Forbes put Elizabeth Holmes's paper net worth at about $4.5 billion that year, then revised it to zero in 2016. Yet according to the SEC, the company told investors it would book more than $100 million of revenue in 2014 when it actually took in a little over $100,000.

What was Theranos supposed to do?

Holmes founded the company in 2003 at age 19, as Real-Time Cures, after filing a patent application for a wearable patch that would monitor blood. The patch proved impractical. The company pivoted to a cartridge and reader: a fingerstick sample in a small "nanotainer" tube, analyzed by a desktop device later called the Edison.

The pitch was cheaper, faster, painless testing. Theranos posted a public price list and said its tests cost no more than half of Medicare reimbursement rates. Revenue would come from fees per test, paid by patients or insurers, with retail partners supplying the footfall. The target was the business of national labs such as Quest Diagnostics and Labcorp, whose patient service centers draw blood by needle.

On paper the market logic was attractive. US lab testing is large, prices are opaque, and patients dislike needles. The problem was physical. Fingerstick capillary blood is variable and contaminated by tissue fluid, and small volumes limit how many assays can be run accurately. Mainstream lab scientists said so from the start.

How did Theranos get to a $9 billion valuation?

Theranos built credibility from everything except data. Its board, assembled largely through George Shultz's network, included Henry Kissinger, William Perry, Sam Nunn, Bill Frist and James Mattis, among others (reported). None were lab experts. Magazine covers followed in 2014.

The late funding rounds came mostly from wealthy families and fund managers, not specialist healthcare investors. Reported investors included Rupert Murdoch ($125 million), the Walton family ($150 million), the DeVos family ($100 million) and Partner Fund Management ($96 million). Trial testimony later showed that a spreadsheet shown to investors in 2013–14 projected about $140 million of revenue in 2014 and about $990 million in 2015. Investors were also shown reports bearing pharmaceutical companies' logos that Theranos had created itself (trial evidence). Major investors did not hire independent lab experts or obtain audited financial statements.

Holmes held about half the equity and majority voting control through super-voting shares. Secrecy was policy: the technology was described as a trade secret, employees signed strict nondisclosure agreements, and the company published almost no peer-reviewed data. In 2015 Stanford's John Ioannidis criticized its "stealth research" in JAMA.

Why did Walgreens and Safeway sign on?

Retailers supplied the distribution Theranos could not build. Walgreens wanted its stores to become health destinations and feared losing the opportunity to CVS. It paid Theranos about $140 million in prepayments, an innovation fee and a convertible note (reported), and opened centers in about 40 stores, mostly around Phoenix. Walgreens's own consultant raised concerns, and Walgreens never got to see the lab (reported).

Safeway went further and got nothing. It reportedly spent about $350 million remodeling hundreds of stores with clinic rooms (WSJ; trial testimony was reported as about $300 million). Theranos repeatedly delayed the launch, a pilot testing Safeway employees produced results that staff questioned, and the partnership ended in November 2015 without a single center opening.

Theranos also used policy. It lobbied for an Arizona law, signed in April 2015, that let consumers order lab tests without a doctor's order, which widened its direct market in the state where its centers were concentrated.

How did Theranos avoid regulators for so long?

The answer is a regulatory gap. Clinical labs are regulated by CMS under the Clinical Laboratory Improvement Amendments of 1988, which set quality standards but do not review whether a test is clinically valid. Tests designed and run inside a single lab were historically left to that regime rather than to the FDA. Theranos ran its tests in its own labs in Newark, California, and later Scottsdale, Arizona, so its device was not reviewed as a product before patients used it.

The FDA did clear one Theranos test, for herpes simplex HSV-1, in July 2015, and the company presented that as validation. It was the only clearance it received. Later that year the FDA found the nanotainer collection tube was an uncleared device (reported). An FDA rule to regulate laboratory-developed tests more broadly was vacated by a federal court on 31 March 2025 (reported), so the gap that Theranos used has not been fully closed. Chapter 20 of the book and the industry study on regulation, compliance and privacy explain how these regimes divide the work.

What did the Wall Street Journal find?

John Carreyrou's article of 15 October 2015 reported that Theranos ran only a small share of its tests on its own devices, used modified commercial machines from Siemens and others, and had accuracy problems. Former employees said the Edison ran around a dozen tests. Whistleblowers Tyler Shultz and Erika Cheung had already complained to regulators.

Theranos chose denial. Its lawyers threatened the Journal and former employees, and Holmes asked Murdoch, the paper's owner and a Theranos investor, to intervene; he declined (reported). The legal intimidation later became trial evidence.

The regulatory sequence that followed was fast. CMS inspected the Newark lab and in January 2016 found that deficiencies posed "immediate jeopardy" to patients, with hematology the worst area. Theranos voided all Edison results from 2014 and 2015 and sent tens of thousands of corrected reports; Walgreens's court filings cite about 1 million voided or corrected results (reported). In July 2016 CMS revoked the Newark lab's certificate and banned Holmes and Balwani from owning or operating a lab for two years. In October Theranos closed its labs and wellness centers and cut about 340 jobs.

What happened to Elizabeth Holmes and Sunny Balwani?

The SEC charged Theranos, Holmes and former president Ramesh "Sunny" Balwani on 14 March 2018, alleging they raised more than $700 million through false statements. Holmes settled without admitting or denying the allegations, paying a $500,000 penalty, returning about 18.9 million shares and accepting a ten-year bar from serving as an officer or director of a public company (reported). A federal grand jury indicted both in June 2018, and the company dissolved that September. Its patents went to Fortress Investment Group, which had lent it $65 million.

On 3 January 2022 a jury convicted Holmes on four counts of defrauding investors and acquitted her on four counts related to patients. She was sentenced to 135 months. Balwani was convicted on all 12 counts, including patient counts, and sentenced to 155 months. A court ordered about $452 million in restitution, and the Ninth Circuit upheld the convictions in 2025 (reported).

Which factor of the Care Equation did Theranos fail?

The Care Equation argues that a healthcare business needs four things at once: care that improves health or lowers cost, a payer with a renewing budget, an affordable route to that payer, and proof strong enough to survive contracting and regulation. The factors multiply, so a zero anywhere gives zero.

Theranos had something resembling the second and third factors. Insurers and patients pay for lab tests every day, and Walgreens supplied a route to them. The first factor, care that works, was never there, and the company replaced the fourth factor, proof, with secrecy, celebrity and a board of statesmen. That is why the book treats Theranos as healthcare's defining story of evidence ignored. In healthcare, proof is not a marketing asset to be produced after scale. It is the condition for being allowed to scale at all.

The case also says something about who checks the proof. Clinicians and lab scientists were skeptical from the beginning, and they were right. The people who were wrong were those who substituted a founder's narrative for a validation study: generalist investors, retail partners and a board without the expertise to ask the right questions. Chapter 4 on value, evidence and outcomes sets out how an expert reads a claim like the one Theranos made.

What are the lessons for healthcare founders and investors?

Four lessons generalize beyond fraud.

Ask to see the lab. Any diagnostics claim can be tested against an established method on the same samples. Theranos never let its most important partner do that. A company that will not share validation data with a paying partner has answered the question.

Price is evidence of cost. Theranos promised tests at half of Medicare rates, which implied a cost breakthrough. In reality it had the costs of a small conventional lab. When the price assumes a breakthrough, check the breakthrough first.

Distribution is not validation. A store footprint, a famous board and magazine covers are reach, not proof. Walgreens supplied the third factor of the equation, and that made the failure of the fourth more harmful because more patients were tested.

Regulatory gaps close. A business model that depends on avoiding review, whether a lab-developed-test loophole or an untested payment rule, is exposed the moment someone looks. The same pattern recurs in very different companies on our list of the biggest healthcare bankruptcies, and in the broader guide to how the healthcare industry makes money.

The takeaway

Theranos failed because it skipped the one step healthcare does not let anyone skip: showing that the product works. It raised hundreds of millions of dollars and won a national retail partner on a story, while running most of its tests on other companies' machines. A regulatory gap delayed the reckoning, and a newspaper and CMS delivered it within a year. For how devices and labs reach the market, get paid and get checked, read Chapter 10, Devices and Diagnostics.