US FDA SaMD Classification System: A Comprehensive Guide

US FDA SaMD Classification System

Written by Pharmadocx Consultants

5 June 2026

The US FDA classifies Software as a Medical Device (SaMD) using a risk-based framework. It places software into four classes (I–IV). Class IV represents the highest risk (life-critical decisions). On the other hand, Class I represents the lowest (supportive, non-serious use). The US FDA SaMD classification plays a central role in the US FDA SaMD registration process. It determines the level of regulatory scrutiny, the type of evidence required, and the appropriate pathway for market authorization. Thus, the classification is the first step in FDA SaMD registration process.

The US FDA SaMD classification system depends on the significance of the information the SaMD provides for healthcare decisions and the seriousness of the condition being managed. By categorizing software into four risk-based classes (I–IV), FDA ensures that oversight is proportional to the potential impact on patient safety. Lower-risk apps may qualify for exemption or 510(k) clearance, while higher-risk, life-critical software requires rigorous Premarket Approval (PMA). This classification guides manufacturers in preparing the right documentation. Basic performance validation is required for Class I to extensive clinical trials and post-market surveillance for Class IV. It also dictates whether a product can follow the 510(k) route (predicate-based), De Novo (novel, low-to-moderate risk), or PMA (high-risk, critical use). Thus, the classification system aligns risk with regulatory expectations and shapes the compliance strategy for safe, effective, and legally marketable software.

US FDA SaMD classification system

Using a risk-based framework, the US FDA classifies SaMD into four classes (I–IV). US FDA SaMD class I to IV are sorted in increasing order of risk level. Higher risk categories require more stringent evidence, validation, and regulatory review than lower risk categories. The US FDA SaMD classification depends on:

  • Healthcare situation/condition: Critical, serious, or non-serious.
  • Significance of information provided: Treat/diagnose, drive clinical management, or inform clinical management.

US FDA SaMD classification

US FDA SaMD classification is risk-based. It ranges from Class I (low impact) to Class IV (life-critical). The classification framework ensures that software delivering medical decisions is regulated proportional to its potential impact on patient safety. We have discussed in details the US FDA SaMD classes.

US FDA SaMD classes

  • Class I (lowest risk): This class covers SaMD that provides information to inform clinical management of non-serious conditions. The software’s role is supportive, offering insights that do not directly drive treatment or diagnosis. FDA considers these tools low-risk because incorrect outputs are unlikely to cause harm. Thus, they often fall under general wellness or preventive health applications. Example: A mobile app that tracks daily hydration levels and reminds users to drink water.
  • Class II (low–moderate risk): SaMD in this class either drives clinical management of non-serious conditions or informs clinical management of serious conditions. These tools influence healthcare decisions but are not the sole determinant of treatment. FDA requires evidence of accuracy and reliability. However, regulatory scrutiny is lighter than higher categories. Risk arises if the software misguides management of chronic but non-life-threatening conditions. Example: A diabetes management app that tracks glucose levels and suggests lifestyle adjustments.
  • Class III (moderate–high risk): This class includes SaMD that treats/diagnoses serious conditions or drives clinical management of critical conditions. The software plays a direct role in guiding interventions where errors could cause significant harm. FDA expects robust clinical validation and quality management systems for these products. These tools often integrate with healthcare workflows and require strong cybersecurity safeguards. Example: Software that adjusts chemotherapy dosage based on patient-specific lab results.
  • Class IV (highest risk): SaMD in this class is used to treat or diagnose critical conditions, where incorrect outputs could be life-threatening. FDA applies the highest level of regulatory oversight, requiring extensive clinical evidence and risk management. These tools often employ advanced algorithms, such as AI-based diagnostic imaging. Lifecycle monitoring, post-market surveillance, and strict validation are mandatory. Example: AI software that detects strokes from CT scans and immediately alerts emergency teams.

US FDA SaMD regulatory pathways

We have presented the different US FDA SaMD regulatory pathways applicable for the different US FDA SaMD classes.

  1. 510(k) premarket notification (Class II, some Class I): The 510(k) pathway is the most common route for SaMD that falls under Class II and certain Class I devices requiring clearance. The 510(k) pathway is used when a SaMD has a predicate device. A predicate device is an already cleared product with similar intended use and technological characteristics. The 510(k) pathway requires demonstrating that the software is substantially equivalent to an already legally marketed predicate device. The FDA reviews safety, effectiveness, and performance data but does not demand full clinical trials unless risk is high. Notably, this pathway is faster and less resource-intensive compared to PMA, thereby making it suitable for moderate-risk SaMD. Example: A diabetes management app that drives clinical management by suggesting insulin dosage adjustments.
  2. De novo classification (novel Class I or II): The De Novo pathway is designed for novel SaMD that has no predicate device but presents low to moderate risk. It allows manufacturers to establish a new device type and classification, thereby creating a regulatory precedent for future submissions. FDA requires evidence of safety and effectiveness, including clinical validation. However, the regulatory burden is lower than PMA. Once granted, the De Novo classification can serve as a predicate for subsequent 510(k) submissions. Example: A new AI-powered dermatology app that identifies skin lesions for early detection of non-critical conditions.
  3. Premarket approval (PMA) (Class III and Class IV): PMA is the most stringent regulatory pathway. It is reserved for Class III and Class IV SaMD that diagnose or treat critical conditions. It requires comprehensive clinical evidence, risk management documentation, and proof of long-term safety and effectiveness. FDA conducts an in-depth review of design controls, cybersecurity, and post-market surveillance plans. Moreover, this pathway is resource-intensive and time-consuming but ensures the highest level of patient safety. Example: AI software that diagnoses strokes from CT scans and directly informs emergency treatment decisions.

An overview of the US FDA SaMD classification and applicable regulatory pathways

SaMD ClassDefinitionExamplesLevel of Regulatory ScrutinyRegulatory Pathway
Category I (Lowest Risk)Provides information to inform clinical management of non-serious conditions.Wellness apps (hydration tracker, fitness reminders).Minimal scrutiny; FDA oversight is light since errors pose negligible risk.Often exempt or 510(k) if predicate exists.
Category II (Low–Moderate Risk)Drives clinical management of non-serious conditions or informs management of serious conditions.Diabetes management app suggesting lifestyle changes; asthma monitoring app.Moderate scrutiny; requires evidence of accuracy and reliability.510(k) if predicate exists, or De Novo if novel.
Category III (Moderate–High Risk)Treats/diagnoses serious conditions or drives management of critical conditions.Chemotherapy dosage adjustment software; cardiac monitoring software.High scrutiny; FDA requires robust clinical validation and QMS.510(k) (if predicate exists) or PMA for higher-risk cases.
Category IV (Highest Risk)Treats/diagnoses critical conditions where errors could be life-threatening.AI stroke detection software from CT scans; sepsis diagnostic algorithms.Highest scrutiny; extensive clinical evidence, post-market surveillance required.PMA (Premarket Approval) is mandatory.

Pharmadocx Consultants: Your trusted support for US FDA SaMD approval

In this blog, we have detailed the US FDA SaMD classification to help you understand the applicable US FDA regulatory pathway. We at Pharmadocx Consultants have extensive expertise in US FDA SaMD regulatory requirements. For any queries or support with US FDA SaMD approval, drop an email at [email protected] or call/Whatsapp on 9996859227.

Looking For a Medical Device or Pharma Consultant?

Blog Categories

Let's Talk!

We'd love to hear from you! Whether you have questions about our pharmaceutical plant setup consultation services or want to discuss a potential project, our team is here to help. Simply fill out the form below, and we'll get back to you as soon as possible. Alternatively, you can reach out to us directly using the phone number or email address listed on this page. We look forward to connecting with you!

Phone / Whatsapp

Address

  • Head Office - Opposite Dewan Mill, Old D.C. Road Sonepat - 131001 Haryana, India
  • Registered Office - Netaji Subhash Place, Delhi, 110034

You May Also Like…

You cannot copy content of this page