Best burglary safes: no safe is burglar proof, so buy the rating

Search for a burglar proof safe and you will be sold steel gauges, bolt counts and the word "impenetrable". None of those are ratings. A rating is a document from a laboratory that describes an attack the safe survived: who attacked it, with what, and for how long. Four of the five safes on this page carry one. The fifth is here because concealment does something no rating measures, and we say so plainly.

We do not test safes. Every figure below is the maker's or the certifying body's published specification, linked at the value it supports and checked 2026-10-19, and each brand carries a dated CPSC recall search on its product card.

Why nothing sold to a household is burglar proof

Every burglary rating is a stopwatch, not a guarantee. UL's UL 1037 standard covers residential security containers, and UL 687 covers burglary-resistant safes: the commercial tier where the TL ratings live. Both describe a tester working for a fixed number of minutes. Pass the clock and the safe is listed. Nobody certifies that a safe can never be opened, because given enough time and the right tool, every one of them can.

That reframes the question usefully. You are not buying invulnerability; you are buying minutes, in a house where the person breaking in wants to be gone fast and is working with whatever was already in the vehicle. Five rated minutes behind an anchored door is a different proposition to a 90 lb box with a "solid steel" sticker on it, and the rated minutes are the only part of that sentence anyone measured.

The four levels you will actually meet

Our TL-15 vs TL-30 vs RSC explainer walks the whole ladder, and UL RSC rating explained covers the residential standard in depth, including the look-alike phrases that imitate it.

How we ranked them

The published rubric at how we rate, with burglary evidence weighted hardest because that is what the search is for:

1. Who issued the rating. A UL or ETL document first. A maker's own claim second. An unattributed adjective last.
2. What the door and bolts are, as published. A 1/2 in steel plate door is a different object to a 12-gauge sheet-metal door with a steel-look paint finish, and the difference is on the spec sheet.
3. Whether the lock is itself listed, and what protects it. A UL-listed lock behind a hard plate with a re-locker is the part of the safe a professional actually goes for.
4. Anchoring, honestly weighted. A rating applies to a safe that cannot be carried away. Bolt holes and included hardware count.
5. Fit for a real home. Weight, footprint and whether you can buy it at all. TL-30 is the best rating here and the worst fit for most houses.

The published evidence, side by side

Read the burglary row and the fire row together and the trade becomes visible. The BF1512 is the only product here with an outside-lab claim in both columns at a size a household can place, and it pays for that with 288 lb around a 15 x 12 x 13 in interior. The Stealth HS4 is the one with an RSC rating you can order on Amazon, and its fire number is a maker claim at 1,200°F with no lab named. The RF6528 wins the burglary column outright and weighs 3,455 lb, which is a conversation with a structural engineer rather than a purchase.

One pattern is worth naming: the safes with the strongest burglary evidence are the ones you cannot buy on Amazon. Three of the five are dealer-sold. That is not a coincidence so much as a market: lab listings cost money, and the brands that pay for them sell through dealers who can explain them.

The B-rate problem, and why a floor safe is still here

Hollon describes the B2500 as a "B-Rated Floor Safe for Foundations". By our own rule that should put it below every listed safe on the page, and on rating evidence alone it is. It is ranked fourth anyway, because a floor safe changes the attack instead of resisting it.

A safe set into a slab has no sides to pry, no back to cut and no way to be wheeled out. Its door sits flush under carpet or a mat, and a burglar working in minutes has to find it first. The published hardware is also genuinely good: a 1/2 in solid steel door, five chrome-plated bolts, a drill-resistant hard plate and a UL-listed Sargent & Greenleaf Group 2 dial with a relocker. What it does not have is a fire rating or a laboratory listing of the container, and installation means cutting concrete, which is not reversible. Our floor safe rankings cover that trade in full, and floor safe vs wall safe compares it with the in-wall option.

The rating assumes the safe is still in the house

Every burglary rating is issued against a safe that is anchored. An unbolted 191 lb home safe on a hard floor is a two-person carry, and a dolly makes it a one-person carry. At that point the attack happens somewhere else, with unlimited time and a cutting wheel, and the five rated minutes are irrelevant.

• Use every hole the maker drilled. The HS4 publishes four bolt-down holes with hardware included; the BF1512 has one recessed anchor hole. Both are there because the maker expects them used.
• Anchor into structure, not into a floating floor. How to bolt down a gun safe covers concrete and wood-framed floors, and the hardware that actually holds in each.
• Check the floor can take it before you buy. Can your floor support a gun safe does the arithmetic for upstairs rooms, which matters once a safe passes about 600 lb.
• Put it where it is awkward to work on. A corner, a closet, bolted with the hinge side against a wall. Where to put a safe covers placement and the trade with fire exposure.

Burglars go for the lock and the drywall, not the plate steel

Thickness arguments assume an attack on the strongest part of the safe. Real forced-entry attempts go for the seams: the lock, the hinge side, the thin back panel, and in a wall safe the drywall around it. That is why the lock specification matters as much as the door.

Two of the picks name a UL-listed lock, which is a listing of the lock rather than the container: the HS4 uses a UL-listed electronic lock behind a hard plate with an internal re-locker, and the B2500 uses a UL-listed S&G Group 2 dial behind a hard plate with an external relocker. A re-locker is the part that fires if someone punches or drills the lock body, dropping an independent bolt that the original mechanism can no longer retract. Electronic vs mechanical locks compares the two entry types on speed, batteries and service, and how hard is it to break into a gun safe walks the attack methods the rated minutes assume.

How to verify a rating claim before you buy

You do not have to take our word for any of this, and you should not have to take the seller's either. Four checks take about five minutes on any product page:

1. Find the standard, not the adjective. "UL RSC", "UL 1037", "RSC Level I", "TL-15" and "TL-30" are standards. "Burglar proof", "bank grade", "maximum security" and "theft resistant" are not.
2. Check whether the maker says it, or only the retailer. This is the single most useful test. Dealer listings routinely attach an RSC claim to safes whose own manufacturer page never mentions one. Open the maker's page for the exact model number.
3. Separate the lock listing from the container listing. "UL-listed lock" is a real and useful claim, and it is not the same as the safe being listed. Both the Stealth HS4 and the Hollon B2500 carry a listed lock; only the HS4 states an RSC rating for the container.
4. Match the model number exactly. Ratings are issued per model, and a brand's flagship listing says nothing about the entry model two rows down the same page. A suffix letter can be the difference between a listed safe and an unlisted one.

If a page passes all four, you have a rating. If it fails the second one, what you have is a retailer's copywriting. Our brand evaluations apply exactly this test at scale: the Cannon review searched every live listing on the maker's own store for UL, RSC, ETL and Intertek, and reports what was and was not there.

Common mistakes

• Treating "burglar proof" or "anti-theft" as a rating. Neither phrase has a standard behind it. Ask which standard and which lab.
• Buying steel gauge instead of a listing. Gauge is one input to a rating, not a substitute for one. Gauge explained converts the numbers to inches.
• Assuming a fire safe resists burglary. Most fire chests are built to insulate, not to resist, and their own pages say nothing about burglary. Fire rating vs burglary rating separates the two.
• Leaving it unbolted because it is heavy. Weight without anchoring is a delay, not a defense.
• Buying TL-30 for a house. At 3,455 lb the question stops being which safe and becomes which room and which floor.
• Forgetting the contents schedule. A rating is also an insurance conversation; ask your insurer what storage a scheduled item requires before you buy.

How to buy a burglary safe

1. Decide what a burglar would take

Cash, jewelry, small electronics and guns are the liquid items, and they set the size. A safe big enough for documents and small enough to carry is the worst outcome of this purchase.

2. Set a rating floor and do not go below it

UL RSC Level I is the sensible residential floor. If a product cannot show you a listing or a maker statement naming the standard, treat it as an access box: good against children and opportunists, not against tools. Our UL RSC roundup lists every listed model we document.

3. Read the lock line as carefully as the door line

Look for a UL-listed lock, a hard plate and a re-locker. Then decide between a dial and a keypad on how you will actually use it, not on which sounds more secure.

4. Plan the anchor before delivery day

Know the floor type, the hardware and the final position before the safe arrives. Moving an anchored safe later is its own project, and moving a safe explains why.

5. Decide whether fire is also in scope

Burglary and fire are separate ratings with separate labs, and very few products carry credible versions of both. If you need both, expect to pay in weight and in dealer purchasing, or to run two containers.

6. Buy the one you will lock

A safe left open because the dial is tedious or the keypad eats batteries protects nothing. Match the entry method to the people using it daily.