Acceptable Quality Level (AQL)

What Is an Acceptable Quality Level (AQL)?

An Acceptable Quality Level (AQL) is the largest share of defective items that a buyer or inspector considers acceptable as the normal quality of a production process. The share of defective items is often called the defect rate. AQL is used in acceptance sampling, where only part of a batch is inspected to decide whether the whole batch is accepted or rejected.

How It Is Used

Manufacturers and buyers often agree on an AQL in a supply contract before production begins. Rather than checking every unit, an inspector pulls a sample from each lot and counts the defective units. The sample size and the maximum number of defects allowed are set by a sampling plan. If the defect count is at or below the allowed number, the lot is accepted. If it is higher, the lot is typically rejected, reworked, or sorted.

Widely used standards, such as ISO 2859-1 and ANSI/ASQ Z1.4, provide tables that link lot size and AQL to a sample size and an acceptance number. The AQL itself is a policy choice made by the parties involved, not a single fixed figure.

Why It Matters to Investors

AQL matters to investors mainly as background for understanding manufacturing quality. Companies that make or buy physical goods may describe quality requirements in annual reports, supplier agreements, or risk disclosures. A pattern of rejected lots can signal higher rework, warranty, or recall costs. An AQL figure on its own, however, says little about a company's overall financial health.

Limitations

AQL is designed so that most lots at the agreed quality level are accepted and most clearly poor lots are rejected. It does not guarantee that every accepted lot is free of defects. The chance that a good lot is rejected anyway is known as producer's risk, and the chance that a poor lot is accepted is known as consumer's risk. AQL also reports how many defects appear in a sample but does not explain why they occur. Finding causes requires tools such as statistical process control and root-cause analysis. The method also works best for counting discrete defects in finished products.