Tamper-Evident Jars for Cannabis: Seals, Bands and Rings

Child resistance keeps a jar hard to open; tamper evidence proves it has not been opened. They are two different requirements — tested differently, supplied differently, and priced differently — and in most US cannabis markets you need both on the same package. This guide explains the four tamper-evident mechanisms that work on glass jars, what they cost to apply, what state regulators actually look for, how a tamper band interacts with a child-resistant closure, and the mistakes that get otherwise good packaging rejected at the dispensary door.

What tamper evidence actually requires

A tamper-evident package has to make the first opening visible. That is the whole requirement, and it comes from the FDA’s tamper-resistant packaging rules for over-the-counter drugs, which US cannabis regulators borrowed almost word for word: the package needs an indicator or barrier to entry that, if broken or missing, provides visible evidence that tampering occurred. Nothing in the rule makes the package harder to open — that is the child-resistant standard under 16 CFR 1700. A package can be trivially easy to open and still perfectly tamper-evident, and vice versa. Cannabis jars in regulated US markets typically need both at once: a CR closure that passes its protocol, and a tamper indicator that breaks visibly on first opening. Conflating the two on a purchase order is the single most common spec error we see.

Four mechanisms that work on glass jars

The breakable cap ring is the default on quality child-resistant glass flower jars because it adds no line step and no separate component: the band is part of the cap moulding, it snaps free the first time a consumer pushes and turns, and it stays parked on the neck as permanent proof of first opening. Shrink bands are the fallback when the cap has no ring of its own, and when you also want batch numbers or a brand mark printed on the seal. Induction seals are the usual choice on glass dropper bottles and on concentrate jars where a hermetic foil barrier also slows terpene loss.

What state regulators actually look for

Most states do not prescribe a specific mechanism. The rule is functional: the package must provide visible evidence of prior opening. In practice, compliance auditors look for three things — the seal exists on every unit, it is intact at point of sale, and it breaks in an obvious way on first opening. California, Colorado, Washington and Massachusetts all use wording close to the FDA OTC model. A handful of states add that the seal must be single-use and non-replaceable: a shrink band printed with a batch code satisfies this, a plain elastic band does not. The state-by-state detail is covered in our guide to US cannabis child-resistant packaging laws; the short version is that the CR closure and the tamper indicator are audited as two separate features, and you should be able to produce a spec sheet for each.

Pairing a band with a CR closure

The tamper-evident component and the child-resistant closure live on the same neck, and they interact in ways that are easy to get wrong. Three rules. First, the band must not block the two motions that make the cap child-resistant — a shrink band applied over the push-and-turn interface can make a cap adult-impossible, which fails the CR test from the other direction. Second, a capped-and-banded unit is a different package in the eyes of the certification protocol; if you band over a CR cap, ask the closure supplier for a compatibility statement and re-run a sample panel on banded units. Third, match the band diameter to the cap skirt, not the glass neck — too loose and it slides in transit, too tight and it splits before anyone opens the jar. When we supply CR flower jars with moulded breakable rings, the certified assembly and the tamper-evident assembly are the same object, so one certificate covers both.

Applying seals at your production scale

Scale decides the mechanism. At 500 to 5,000 units a month, a heat gun and pre-cut bands are fine: an operator bands and shrinks 8 to 12 jars a minute, and tooling costs under $100. From 5,000 to 50,000 units, a bench-top heat tunnel pays for itself within months — throughput rises to 40–60 units a minute with far more consistent shrink, which matters because inconsistent shrink is the top cause of bands splitting in transit. Above that, the band applicator and tunnel go inline with the capper. If you want no application step at all, specify a closure with a moulded breakable ring and the tamper evidence arrives free with the cap. Induction sits in between: handheld sealers are cheap, but the foil liner must match the jar mouth diameter and the product — essential oils and high-terpene extracts attack some adhesives, so request a liner compatibility test against your actual formula, not a generic one.

Testing before you commit to a container load

Run three tests on sealed samples before mass production. Transit: pack the jars exactly as they will ship — carton, divider, pallet — and drop the carton on a corner from one metre; a band that splits in the box fails twice, at compliance and at returns. Opening: have ten adults open the sealed jar; if any of them defeats the band without visible damage, the band is decorative. Climate: bands applied too tightly can split in a hot truck in July; PET bands are more temperature-stable than PVC and are the safer choice for freight routed through hot climates. Any competent packaging supplier runs these on samples before you pay for tooling; ours does it as a standard pre-production step, and we send the photos with the sample report.

Common mistakes that sink otherwise good packaging

• Writing “tamper-evident” on the PO without naming the mechanism — you will get whatever cap is on the shelf
• Banding over the CR interface and assuming the CR certificate still applies
• Specifying PVC bands for shipments routed through hot climates
• Printing batch codes on the band, then finding the ink offsets onto the glass
• Testing the seal on a bench but never inside a packed, palletised carton
• Forgetting that any change of cap supplier invalidates both the CR and the tamper-evident validation