How we test metal detectors

The five criteria

1. Depth on realistic targets. Not a laboratory air test — buried coins and jewelry-sized targets at measured depths in real soil, checked across at least two ground types. Advertised depth figures are ignored entirely.

2. Separation in trash. How well the machine picks a good target out from beside a nail or a bottle cap. This is where most real-world finds are won or lost, and it's where cheap machines fall apart fastest.

3. Ergonomics over a long hunt. Weight matters, but balance matters more. A detector is judged on how it feels at hour three, not at unboxing — including grip, armrest, and whether the screen is readable at dawn and dusk.

4. Build quality and support. Coil-ear strength, shaft wobble, waterproof claims taken at their stated rating, and the practical question of whether you can get a spare coil in under two weeks.

5. Value against the alternatives. A rating is always relative to what else exists at the price. A very good $200 machine can out-rate a merely adequate $900 one, and often does.

What the score means

Ratings run 0–5 in half-point steps and are shown as a target ID readout, because that's the number a detector uses to score a find.

4.5–5 — Genuinely excellent for its price class; the machine to beat. 4 — Very good, with a real limitation that's clearly stated. 3–3.5 — Works, does its job, clearly outclassed by better options nearby. Below 3 — Only recommendable in a narrow circumstance, and the review says which.

A 3.0 is not a failing grade. The Bounty Hunter Tracker IV scores a 3 and it's still the right buy for someone testing the hobby — the rating reflects the machine, the verdict reflects the buyer.

What we deliberately ignore

Manufacturer depth claims. They're best-case laboratory figures on large targets in clean soil, and they bear almost no relationship to what happens in mineralized ground.

Frequency counts and mode counts. More is not better. A well-chosen 10 kHz machine beats a badly matched multi-frequency one in the same dirt every time.

Marketing language. "Professional grade" and "advanced technology" are not specs and don't enter a rating.

Independence

Manufacturers do not review, approve, or see articles before publication. No payment has ever been accepted in exchange for a rating, a placement, or a recommendation, and none will be.

Affiliate commissions are how this site is funded. They do not influence ratings — several of the top picks here are the cheapest machine in their category, which is the opposite of what a commission-driven site would do.

Corrections and updates

Reviews are revisited when a machine is superseded, when a firmware update materially changes its behavior, or when a reliability pattern emerges in owner reports. Every page shows its last-updated date.

If something here is wrong, tell us and it gets fixed — with the change noted rather than quietly edited.