USP <790> Visible Particulates in Injections is a mandatory chapter. It requires injections to be essentially free of visible particulates and inspected 100% during manufacturing. USP <1790> Visual Inspection of Injections is a guidance chapter. It explains how to build the inspection programme: methods, lighting, inspector qualification, defect classification and trending.
Every injectable product has to be free of visible particles, and every pharmaceutical quality team has to prove it. In the United States Pharmacopeia, that obligation sits in two chapters that are often confused with each other: USP <790> and USP <1790>. One tells you what you must achieve. The other tells you how to achieve it.
This guide explains what each chapter covers, where the requirements differ, what inspection conditions they describe, and what an inspection programme needs in order to stand up to an audit.
Why There Are Two Chapters
USP uses its chapter numbers to signal status. Chapters numbered below <1000> are requirements, and they are enforceable when a product claims compliance with a USP monograph. Chapters numbered above <1000> are informational, and they give recommended practice rather than binding rules.
That single difference explains why <790> is short and <1790> is long. <790> sets the standard that the product must meet. <1790> gathers the accumulated industry practice that helps manufacturers meet it consistently.
USP <790> vs USP <1790> at a glance
| Point | USP <790> | USP <1790> |
| Status | Mandatory requirement | Informational guidance |
| Main question it answers | What must the product meet? | How do we inspect to meet it? |
| Scope | Visible particulates in injections and implanted drug products | The whole visual inspection programme, including containers and cosmetic defects |
| Length and detail | Short and prescriptive | Long, with background and examples |
| Typical use | Quoted in specifications and audits | Used to design SOPs, training and qualification |
What USP <790> Requires
The core requirement of <790> is that injections must be essentially free of visible particulates. The chapter also sets out how this is demonstrated during manufacture and how a lot already released can be assessed later.
• 100% inspection during manufacturing. Every filled unit of the lot is inspected, not a sample. This applies whether the inspection is manual, semi-automated or fully automated.
• A described inspection method. Units are inspected against black and white backgrounds under controlled lighting for a defined viewing time.
• Defect removal. Units found with visible particulates are rejected and removed from the lot.
• Assessment of released lots. For product that has already completed 100% inspection, the chapter describes assessing a sample using an accepted statistical sampling plan with a defined acceptance number, rather than repeating a full inspection.
• Scope beyond vials. The requirement covers injections in all container types, including ampoules, prefilled syringes and cartridges, as well as implanted drug products.
What “essentially free” actually means
It does not mean a guaranteed zero. Visual inspection is a probabilistic test, and detection depends on particle size, contrast, container clarity and the inspector. “Essentially free” means the lot has been 100% inspected by a qualified, validated process and that rejects have been removed, so the probability of a particle-bearing unit remaining is very low.
The Inspection Conditions the Chapters Describe
Both chapters describe a manual inspection carried out at a controlled station. The conditions below are the reference point that most inspection SOPs are built on, and they are also the conditions in which inspector qualification should be carried out.
Reference conditions for manual visual inspection
| Element | What is described |
| Backgrounds | A matt black panel and a non-glare white panel, so both light and dark particles can be seen |
| Illumination | Roughly 2,000 to 3,750 lux at the point of inspection for clear glass containers |
| Difficult containers | Higher illumination for plastic, amber or tinted containers, where light transmission is lower |
| Viewing time | About 10 seconds per container, split between the black and the white background |
| Handling | Gentle swirling or inversion to suspend particles, without creating air bubbles that look like particles |
| Container preparation | The outside surface cleaned and dried so external dust is not mistaken for product particles |
| Inspector comfort | Controlled inspection rate, scheduled breaks and rotation to limit fatigue |
What USP <1790> Adds
<1790> takes the same activity and turns it into a managed programme. It is the chapter to read when you are writing an SOP, designing training, choosing between manual and automated inspection, or defending your approach in an audit.
1. Inspector qualification and requalification
<1790> describes qualifying inspectors against a test set of units that contains both defective and acceptable containers, covering a range of defect types and detection difficulty. Each inspector’s results are compared against a known reference, and performance is measured rather than assumed. This is the approach most laboratories apply through a structured inspector qualification process, and the chapter also supports periodic requalification, which we cover in our post on how often inspectors should be requalified.
2. Defect classification and acceptance limits
The chapter recommends grouping defects into categories by the risk they carry, and applying different acceptance limits to each category. Critical defects, which affect patient safety or container integrity, carry the tightest limits. The example limits below are the pattern used across the industry, but the exact figures must be set in your own quality system and justified.
Typical defect categories used in injectable inspection
| Category | Examples | How it is treated |
| Critical | Visible particles, cracked containers, leaking or missing seals | Tightest acceptance limits; these defects drive patient risk |
| Major | Under or over fill, damaged stopper, serious cosmetic faults that affect use | Intermediate limits |
| Minor | Small cosmetic marks, minor label or appearance issues | Widest limits |
3. Choice of inspection method
Manual, semi-automated and fully automated inspection are all acceptable. <1790> explains what each suits and stresses that an automated system must be shown to perform at least as well as the qualified manual process, using the same kind of test set. That comparison is the reason a well-characterised defect kit matters as much for machine qualification as it does for people.
4. Products that are difficult to inspect
Lyophilized products, suspensions, emulsions, coloured solutions, amber glass and prefilled syringes are all harder to inspect than a clear solution in clear glass. <1790> discusses adapting the method for these, for example by using different lighting or specific handling. Our guide on inspecting lyophilized vials looks at the cake-related defects that complicate this further.
5. Trending and knowledge management
The chapter also expects reject data to be collected, trended and investigated, so the inspection programme feeds back into the process. A rising reject rate for one defect type is a process signal, not just a quality result.
How These Chapters Fit With Other Requirements
Visual inspection is only one part of the picture, and auditors will look at how the pieces connect. Visible particulates sit under <790> and <1790>, while smaller particles fall under the subvisible particulate chapters, and container closure integrity is handled separately.
Where each requirement sits
| Requirement | Where it is covered |
| Visible particulates in injections | USP <790> (mandatory) and USP <1790> (guidance) |
| Subvisible particulate matter | The USP subvisible particulate matter chapters, tested by instrument rather than by eye |
| Container closure integrity | USP <1207>, explained in our post on what USP 1207 means for CCIT |
| EU expectations | EU GMP Annex 1, which requires 100% inspection and qualified, periodically re-checked inspectors |
| European Pharmacopoeia | Ph. Eur. 2.9.20, the corresponding European test for particulate contamination |
What a Compliant Inspection Programme Looks Like
Putting both chapters together, a programme that holds up in an audit usually has the following elements in place and documented.
1. A written SOP that defines the inspection station, lighting, backgrounds, viewing time and inspection rate.
2. Calibrated light measurement at the inspection point, with records.
3. A defect library or test set that reflects the defects your own products actually produce.
4. Initial qualification of every inspector against that test set, with recorded results.
5. Scheduled requalification, with a defined pass criterion and a plan for inspectors who fail.
6. Vision testing for inspectors, repeated at a defined interval.
7. A defect classification list linking each defect to a category and an acceptance limit.
8. 100% inspection records for every lot, with reject counts by defect type.
9. Trending of reject data, with investigation triggers.
10. Qualification data for any automated inspection machine, compared against the manual process.
Common Gaps Auditors Find
Most findings in this area are not about the chapters being misunderstood. They are about the programme drifting away from what the SOP says.
• Light levels measured once at installation and never checked again.
• A test set that contains only easy, obvious defects, so every inspector passes.
• Inspector qualification records that show a pass but not the underlying results.
• Requalification overdue, or carried out only when someone remembers.
• Inspection rate in practice far faster than the rate the SOP and qualification assumed.
• Reject data collected but never trended or investigated.
Frequently Asked Questions
1. Do USP chapters apply to a manufacturer in India?
They apply when a product is supplied to the United States market or when a monograph or contract requires USP compliance, and US regulators expect the requirements to be met at the manufacturing site wherever it is located. Many Indian manufacturers also adopt them as internal standards because customers and auditors in other markets recognise them.
2. Does an automated inspection machine remove the need for qualified human inspectors?
No. Machines still need qualified people to handle rejects, to carry out the manual re-inspection of machine rejects and to run the comparison studies that keep the machine qualified. The machine itself must also be qualified against a characterised set of defective and acceptable units.
3. How many units should a qualification test set contain?
The chapter does not fix a single number. A set has to contain enough defective and acceptable units to cover the defect types you expect and to span a range of detection difficulty, so that the result distinguishes between strong and weak inspectors. Sets used in practice usually run to a few hundred units in total, and the composition should be justified in your own protocol.
4. Can we reuse the same test set every year?
A set can be reused if it is stored correctly, checked for stability of the defects and re-certified at a defined interval. The risk with reusing the same units too often is that inspectors begin to recognise individual containers rather than the defects, so many sites rotate the arrangement or maintain more than one set.
Building an Inspection Programme That Meets the Standard
The two chapters ultimately ask for the same thing from two directions: a process that reliably finds visible particulates, run by people whose ability to find them has been measured. The lighting, the backgrounds and the viewing time are easy to set up. The harder part is proving, with records, that the people and the machines doing the inspection perform to a known standard.
Mahant Pharma Advances manufactures Knapp test kits, visual inspection challenge sets and CCIT challenge sets used by pharmaceutical laboratories to qualify inspectors and inspection machines against characterised defects.