Palo Alto Health Labs All services
Tool · Regulatory

Is my health software a medical device?

An investor asks "is this a device?" and you need a straight answer. Describe one feature and get a draft read, with the FDA rule behind each flag, in about 3 minutes.

Format
Interactive checker
Time
About 3 minutes
Sources
FDA guidance
Checked
October 27, 2026

New to the rules? Read the five-question FDA decision guide first.

Answer for one software function at a time. FDA classifies each software function separately (FDA Step 1).

Your answers stay in your browser. Nothing is saved or sent.

Draft and educational only. Not legal or regulatory advice, and not an FDA determination. Palo Alto Health Labs is not regulatory counsel or a filer of record.

What is this feature meant to do? Go by what your website, app-store listing, deck and in-app text say, not only by what the code does.

FDA judges software by its intended use, and your claims set that. Rewording one sentence of marketing can move a feature in or out of device territory.

Source: FDA Step 1 (201(h) definition: "intended for use in the diagnosis of disease or other conditions, or in the cure, mitigation, treatment, or prevention of disease"); General Wellness guidance, Jan 2026, two categories of wellness intended uses.

Do your labeling, ads, sales deck, UI or features include any of these? Select all that apply.

Check every place a customer or investor could see it, including the deck. Even one of these usually takes a product out of general wellness.

Source: General Wellness guidance (Jan 2026), list of 5 features that mean "not general wellness products".

Does the product measure or estimate signals from the body?

If you only estimate things like heart rate variability from a non-invasive sensor, you can still be in wellness territory. What you show and claim about those numbers decides it, which the next question checks.

Source: General Wellness guidance (Jan 2026), section on non-invasive sensing.

Does it show numbers that look like clinical readings, such as blood pressure in mmHg or SpO₂ %?

Trends and scores are treated differently from numbers a clinician would read as a measurement. See the decision guide for how this plays out.

Source: General Wellness guidance (Jan 2026): "do not include values that mimic those used clinically unless validated (e.g. manufacturer testing, peer-reviewed clinical literature)".

Who reads the output and acts on it?

Source: CDS guidance (Jan 2026), Criterion 3 (recommendations "to a health care professional"). The guidance also says existing digital health policies "continue to apply to software functions … intended for use by patients or caregivers" (FDA CDS guidance page).

Does this feature acquire, process or analyze a medical image, a signal from an in-vitro diagnostic, or a physiological signal or pattern (for example ECG, EEG, CT or a continuous sensor waveform)?

If yes, the feature can't qualify as non-device clinical decision support, whoever uses it. Working only from recorded information like notes, labs or guidelines is a “no.”

Source: CDS guidance, Criterion 1, section 520(o)(1)(E).

What does it give the clinician?

How specific and how urgent the output is matters as much as what it says. The decision guide explains why.

Source: CDS guidance, Criteria 2 and 3. The guidance says FDA considers "the (a) level of software automation and (b) time-critical nature of the HCP's decision making". For software that otherwise meets the criteria but gives "only one recommendation that is clinically appropriate", FDA "intends to exercise enforcement discretion" (CDS guidance, Jan 2026).

Can the clinician see and check the basis for the output (the inputs used, the logic or sources, and known limits), so they aren't relying mainly on the software?

A score or flag with no visible reasoning counts as “no.” The decision guide covers this criterion in more detail.

Source: CDS guidance, Criterion 4.

For patients, does it only help them track or manage a condition, without giving specific treatment suggestions?

Logging, reminders and general education count as yes; interpreting data to flag a condition counts as no. See the decision guide for FDA's examples.

Source: Policy for Device Software Functions and Mobile Medical Applications (Sept 2022), §V.B: enforcement discretion for functions that "Help patients … self-manage their disease or conditions without providing specific treatment or treatment suggestions; or 2. Automate simple tasks for health care professionals."

Do you know of a device already legally on the US market with the same intended use and similar technology?

A best guess is fine, since this only shapes the pathway hint, not the result. The decision guide outlines the pathways.

Source: FDA 510(k) page (substantial equivalence to a legally marketed "predicate"); De Novo page (novel devices with "no legally marketed predicate device", class I or II).

Related

Questions

Does FDA classify a whole app or each software function?

Each software function. FDA's Step 1 page says to repeat the assessment for each software function in a product.

Can a wellness wearable show blood pressure or SpO2 values?

Under FDA's January 2026 General Wellness guidance, non-invasive wellness products should not display values that mimic clinical ones unless validated, and must avoid disease references, clinical alerts, treatment guidance and clinical-grade claims.

Is 510(k) or De Novo the right pathway for my software?

A 510(k) shows substantial equivalence to a legally marketed predicate device. De Novo is for novel low-to-moderate-risk devices with no legally marketed predicate. A Pre-Submission can get FDA feedback first.