Car anti-theft devices that work are the ones matched to the right job. Every gadget in this category does one of three things: prevent the theft outright, delay it long enough that the thief gives up, or do nothing to stop it but help you recover the car and prove what happened afterward. Buying a delay device and expecting prevention, or a recovery device and expecting either, is where most disappointment with these products actually starts.
This guide sorts the whole category by that one distinction, states each class's honest limits with a named source where one exists, and points out the one class of device you should personally test before trusting it. Nothing here is sold, and no brand is being recommended.
Prevent, delay, recover: the three jobs a device can do
Every anti-theft gadget fits into one of three jobs, and mixing them up is the root of almost every complaint about a product "not working."
Prevent means the theft attempt technically fails: a signal-blocking pouch stops a relay attack from ever reaching the fob, an aftermarket immobilizer breaks a circuit the thief cannot bridge. Delay means the theft is still technically possible but now takes longer than most thieves are willing to spend in the open: a steering wheel lock, a wheel clamp. Recover or evidence means the device does nothing to stop the theft but changes what happens after it: a hidden GPS tracker, a dash cam facing the cabin. None of these substitutes for either of the other two. A tracker will not stop a car from being driven away, and a steering wheel lock will not tell you where it went.
Relay theft and why keyless cars changed the problem
Keyless entry and push-button start systems work by listening for a fob's short-range radio signal and unlocking or starting the car when it answers. A relay attack exploits that trust: one device near the house captures the fob's dormant signal, a second device near the car relays it, and the car behaves as though the fob is standing next to it, even though the fob and the owner never left the hallway.
The National Insurance Crime Bureau tested a commercially available relay attack device in 2016 across 35 vehicle makes and models. It opened 19 of them (54 percent), started and drove away with 18 (51 percent), and in 12 cases (34 percent) restarted a car that had already been switched off. NICB's own conclusion at the time was blunt: the devices work, not on every car, but on enough of them that a thief can target and steal them with relative ease. Vehicle security and attack hardware have both moved on since 2016, but the underlying weakness, a system that trusts whichever signal answers first, is specific to keyless cars and is why they carry a prevention step that older keyed vehicles never needed. The full mechanics of the attack are covered at relay theft, explained for keyless car owners.
NICB's separate annual theft tally gives the wider context: it reported roughly 850,000 vehicle thefts in the US in 2024, a 17 percent drop from 2023 and the largest one-year decline in decades, with the decline continuing through 2025. Even with the total falling, the count is still large enough that which device class an owner picks, and whether it is actually tested to work, keeps mattering.
Signal blocking, and why you must test yours
A Faraday pouch or box works by physically shielding the fob inside a layer of conductive material, creating a cage the fob's radio signal cannot pass through, so there is nothing for a relay device to capture. Done correctly, it is a genuine prevent, not a delay.
The catch is construction quality varies across the market: mesh density, seam continuity, and how well the closure seals all affect whether shielding actually holds. Independent publications that test signal-blocking bags, RECOIL OFFGRID among them, have found meaningfully different results depending on material and build even among products marketed for the same purpose. None of that is visible from a product photo or a listing description. The only way to know a specific pouch or box works is to test the specific one you own: seal the fob inside, stand next to the car, and try both the remote unlock button and a push-to-start attempt. If either responds, the shielding or the seal has failed for that unit. The full step-by-step version of this test, and what a fail actually means, is at Faraday pouches for key fobs: do they work, and how to test yours. For the choice between a pouch, a rigid box, and simply putting the fob into its own sleep mode, see signal-blocking box, pouch, or just turning the fob off.
Physical deterrents and defeat time
Steering wheel locks, wheel clamps, and tire boots are the classic delay category: visible, physical, and intended to make the job take longer than a thief will spend in the open. Their honest limit is that "longer" is not a fixed number, it depends entirely on the build.
Car and Driver's Gear Team tested five steering wheel locks side by side. One held up against every method the testers tried: hand-shaking, hammering, drilling, prying. Two others were defeated with bolt cutters once the testers found the weak point, a thin connector cable in one case, and the lock's own shaft in another, after a single hammer strike loosened it enough to get the cutters onto it. That is the honest range for this single device class: from unbreakable in testing to broken in under a minute, sold at similar price points, with no way to tell the two apart from a listing photo. A visible lock still changes a thief's calculation just by being seen, since it signals the car will take longer than the one parked next to it, but "steering wheel lock" alone is not a spec, and the underlying bar and lock-body construction is what actually decides defeat time. The deterrent-value question in full, along with what to look for, is covered at steering wheel locks: deterrent value vs. defeat time.
Immobilizers and the OBD port as a way in
A factory immobilizer already requires the correct chip inside the key before the engine will start, which is why hot-wiring a modern car with a screwdriver mostly stopped working decades ago. The vulnerability that replaced it is physical rather than electronic: if a thief gets access to the OBD-II diagnostic port, commonly located under the dash, commercially available diagnostic tools can read the car's immobilizer data and program a blank key to work with it, an attack path that never needs the original fob at all.
An OBD port lock, either a physical case that encloses the port or an electronic lock that works with an aftermarket immobilizer, closes that specific path. It is a prevent for that one attack, but it does not stop a thief from breaking a window or picking a door lock to reach the port in the first place; that is a separate gap it does not cover. The full comparison of physical vs. electronic OBD locks and what each actually blocks is at OBD port locks and aftermarket immobilizers.
Trackers, and what they can and cannot do
A GPS tracker does nothing to prevent or delay a theft. It sits entirely in the recover or evidence category, and treating it as a deterrent is the most common mismatch in this whole category. A cellular tracker reports its position over a data connection, sometimes with data included in the purchase price rather than a recurring subscription; a Bluetooth tracker instead relies on nearby phones in a crowdsourced network and has a much shorter effective range, which suits a car that stays near familiar places better than one that could be driven anywhere.
Either way, a tracker only helps if it has power, a signal, and someone actually watching for an alert once the car moves unexpectedly; it changes nothing about whether the theft happens. Where in the car it should physically live, how it actually reports a position, and the legal line on tracking a vehicle someone else drives, all deserve their own answers rather than a guess here: see how a car GPS tracker actually reports position, GPS trackers without a monthly fee: what you give up, where a tracker can live in a car, and why power decides it, and the legal line on tracking a vehicle you don't own. Laws on vehicle tracking vary by state and change; confirm your own state's current rules before installing one on a car that is not solely yours.
What the vehicle and the parking spot decide
The same device performs a completely different job depending on the car it is protecting. A keyless, push-button-start vehicle needs a relay-theft answer that an older car with a physical key and no proximity fob simply does not need. A car that already came with a factory-fit alarm and immobilizer is starting from a different baseline than one that did not.
Parking location changes the outcome as much as, and often more than, the gadget chosen. A locked garage removes most opportunistic theft on its own; a well-lit driveway visible from the street is a meaningfully different risk than an unlit spot around the side of a house; a car parked somewhere with no security camera coverage loses the evidentiary benefit a dash cam or doorbell camera would otherwise provide. None of that shows up on a product spec sheet, and it is worth thinking through before comparing devices at all.
Layering: a realistic order to add things
No single device covers all three jobs, which is why every credible source on this topic converges on layering rather than picking one gadget and stopping. A reasonable order to add them:
First, close whatever gap the car's own factory security does not already cover: a signal-blocking solution for a keyless car vulnerable to relay theft, or an OBD port lock if the car has no aftermarket immobilizer behind that port. Second, add a visible delay device that changes a thief's on-the-spot calculation, since a car that visibly takes longer is a less attractive target next to one that does not. Third, add a recovery layer last: it will not have stopped anything, but it changes the odds of getting the car back and gives police and an insurer something concrete to work with. Skipping straight to a tracker while leaving the prevention gap open is the most common ordering mistake, because it is the layer that does the least to actually stop a theft in progress.
What is mostly theatre
Some products in this category sell the appearance of security rather than a change in what a thief has to do. A loud alarm that bystanders have learned to tune out, a blinking red "security system" LED with nothing behind it, a decal claiming a tracker or immobilizer is installed when none is, an unbranded signal blocker sold with no test data behind it at all: these depend entirely on convincing a thief something is present, rather than on actually changing the physical or electronic obstacle in front of them.
That is not to say deterrence has no value, a visible device does change a thief's calculation. But a real steering wheel lock, plainly visible through the windshield, beats a decal every time, because there is nothing left to bluff once the thief is standing at the car. The honest test for any device in this category is whether it still does its job if the thief already suspects it might be fake.
A comparison table of device class against the job it does
| Device class | Job | Honest limit |
|---|---|---|
| Faraday pouch or box | Prevent (blocks relay signal) | Only works if fully sealed, and construction quality varies; test your own unit |
| Steering wheel lock | Delay | Defeat time varies widely by build, from under a minute to holding against every method tried, per independent testing |
| OBD port lock | Prevent (blocks reprogramming path) | Doesn't stop entry itself; the break-in or lock-pick path still needs closing separately |
| Aftermarket immobilizer | Prevent | Adds a barrier beyond factory security; install quality decides how solid it actually is |
| GPS tracker | Recover / evidence | Does not stop the theft; needs power, signal, and someone watching the alert |
| Loud alarm | Marginal delay / deterrent | Widely tuned out by bystanders in practice; easy to overrate |
| Wheel clamp or tire boot | Delay | Visible and heavy; works mainly on a parked, unattended vehicle |
Deep dives: the rest of this cluster
This guide is the overview. Each of these goes deeper on one piece of it:
- Faraday pouches for key fobs: do they work, and how to test yours, the actual test procedure
- Relay theft, explained for keyless car owners, the attack mechanism in full
- Signal-blocking box, pouch, or just turning the fob off, the three options compared
- Steering wheel locks: deterrent value vs. defeat time, what to look for beyond the category name
- GPS trackers without a monthly fee: what you give up, the subscription-free tradeoffs
- How a car GPS tracker actually reports position, cellular vs. Bluetooth explained
- The legal line on tracking a vehicle you don't own, where consent matters
- Where a tracker can live in a car, and why power decides it, the placement question
- OBD port locks and aftermarket immobilizers, the reprogramming path closed
FAQ
What is the single most effective car anti-theft device?
There isn't one, because no device covers all three jobs. The most effective approach is layering: close the specific gap your car's own factory security leaves open, add a visible delay device, then consider a recovery layer.
Do Faraday pouches actually stop keyless car theft?
A properly sealed, well-built one blocks the fob's signal and prevents the relay attack that keyless theft depends on. Construction quality varies across the market, so test your own pouch by sealing the fob inside and trying to unlock and start the car from outside; if it responds, the pouch has failed for that unit.
Are steering wheel locks worth it if a thief can just cut them off?
Defeat time varies a great deal by build; independent testing has found locks defeated in under a minute and others that resisted hammering, drilling, and prying. A visible lock still changes a thief's on-the-spot calculation, but not every lock offers the same resistance.
Can a GPS tracker prevent my car from being stolen?
No. A tracker does nothing to stop a theft in progress; it only helps after the fact, by reporting the car's location if it has power, a signal, and someone watching for the alert.
Is it legal to install a GPS tracker on my own car?
Generally yes for a vehicle you fully own, but the rules change once someone else drives or co-owns the car, and laws vary by state. Confirm your own state's current rules before installing one.
What actually gets into a car through the OBD port?
A thief who reaches the OBD-II port with a commercially available diagnostic tool can read the car's immobilizer data and program a blank key, without ever needing the original fob. Reaching the port still requires getting into the car first, which an OBD lock does not prevent on its own.
Do car alarms still deter thieves?
A visible, functioning alarm adds some deterrent value, but a loud alarm alone is widely tuned out by bystanders in practice and shouldn't be relied on as the only layer of protection.
Vehicle theft methods, security technology, and vehicle-tracking laws all change over time and vary by state. Confirm current rules with your own state's authority, and check your specific vehicle's factory security features against its owner's manual before adding an aftermarket device.