Bottle QC Checklist: The Seven Tests Every Shipment Should Pass

Most bottle defects do not appear at the factory. They appear three months later, in a warm warehouse, on a filled bottle that has been sitting on a shelf. By then the discussion is no longer about quality control — it is about who pays.

These are the seven checks that actually predict whether that happens. Each one is a standard test; none of them requires a laboratory.

1. Dimensions and Weight

The starting point, and the one most often done casually. What matters:

• Neck finish. Thread and sealing surface measured with a go/no-go gauge, not eyeballed. This single dimension decides whether your closure leaks.
• Wall thickness. Measured at three points around and three heights up the body. The figure that matters is the minimum, not the average — an average hides a thin panel.
• Overall height and diameter. Within about ±0.3 mm for blow-moulded plastic, tighter for the neck interface.
• Unit weight. Kept inside ±2–3% across a batch. Wide weight scatter is the signature of an unstable parison, which always comes with uneven walls.

2. Leak Test on the Filled Pack

Filled with the actual formulation, or a surrogate matched for viscosity, then sealed with the production closure, inverted, and held at 40 °C for 24 hours. Acceptable weight loss is no more than 0.2% of fill weight.

For a pump or lotion dispenser the test continues: 20 actuations in the inverted position, with output per stroke staying within ±10% of specification. A pump that delivers correctly upright and dribbles upside down has failed.

3. Drop Test

Filled, conditioned at 5 °C for four hours, then dropped from 1.2 m onto a concrete surface in three orientations — base, side, corner. For plastic packaging the pass criteria are: no cracks, the closure stays sealed, and a pump does not come apart.

The target is 100% pass on 30 samples. Photographs of one surviving bottle are not a result.

4. Thermal Cycling

Five cycles of 24 hours each: 4 hours at 50 °C, 4 hours at room temperature, 4 hours at 5 °C, then 12 hours at room temperature. Afterwards inspect for leakage, deformation, closure loosening and product discolouration.

This test finds the problems that a single-temperature leak test misses — a bottle that seals hot and seeps cold, or a label that lifts after two cycles.

5. Top Load Strength

Vertical compression resistance, measured on an empty bottle. It determines how high your pallet can be stacked and whether a case can survive being bottom of a stack in a fulfilment centre. Under-designed top load is the most common reason bottles arrive oval instead of round.

6. Torque and Closure Retention

Application and removal torque, typically in the range of 1.2–2.5 N·m for cosmetic closures. Too low and the seal fails in transit; too high and the consumer cannot open it. Where tamper-evidence or child resistance applies, this test is also what proves the feature works.

7. Compatibility and Accelerated Stability

Filled packs stored at 40 °C ±2 °C and 75% relative humidity for three to six months, then inspected for dimensional change, discolouration, surface deformation and weight loss through permeation. This is the test that catches incompatibility between formula and container — swelling, crazing, or a soft bottle that was rigid when it left the factory.

How Many to Inspect

Sampling follows ISO 2859-1 rather than a fixed percentage. A workable defect classification for cosmetic packaging:

What to Ask For

• The dimensional report from your production run — not a generic specification sheet.
• Drop and leak test results with sample sizes and temperatures stated.
• Photographs of the actual shipment at inspection, not catalogue images.
• Confirmation of who performed the inspection and against which standard.

The Short Version

Seven tests: dimensions and weight, leak on the filled pack, drop at cold temperature, thermal cycling, top load, torque, and compatibility. None requires exotic equipment. What they require is that someone actually runs them, records the numbers, and is willing to show you the records — which is the real difference between a factory and a trader.