How to Spot Fake Gold — The Complete Guide

Gold bullion counterfeiting is a real and growing risk. As gold prices have risen over the past two decades, the financial incentive to produce convincing fakes has grown proportionally. Modern counterfeiters use industrial-grade equipment and materials that can fool even experienced dealers.

The most dangerous fakes use tungsten cores plated with genuine gold. Because tungsten’s density (19.25 g/cm³) is nearly identical to gold’s (19.30 g/cm³), these fakes pass the most commonly used tests: weight checks, water displacement, and even X-ray fluorescence surface analysis.

Whether you’re buying from a dealer, at a coin show, or through a peer-to-peer marketplace, knowing how to verify authenticity protects your investment and gives you confidence in every transaction.

Start every authentication by examining the coin carefully under good lighting. Look for sharp, well-defined details in the design — the fine detail of Britannia’s shield and robes, the eagle feathers on an American Gold Eagle, or the edge reeding on a Krugerrand. Genuine coins are struck with enormous pressure using hardened steel dies, producing crisp, consistent detail.

Check the colour under natural light. Pure 24-carat gold (used in Maple Leafs, Philharmonics, and Kangaroos) has a distinctive warm yellow. Alloyed 22-carat gold (Krugerrands, Sovereigns, Eagles) has a slightly redder tone due to copper content. A coin that looks too pale, too orange, or has an uneven colour may be suspect.

Examine the edge. Many coins have reeded (ridged) edges or specific lettering. Count the reeds if possible — they should be uniform and evenly spaced. Poorly made counterfeits often show irregular edges, seam lines, or slightly rounded reeds.

Every gold coin has precisely specified mass, diameter, and thickness values published by the issuing mint. A genuine 1 oz Krugerrand weighs exactly 33.93 grams, measures 32.77 mm in diameter, and is 2.84 mm thick. Deviations beyond manufacturing tolerances of ±0.05 g (weight) and ±0.15 mm (diameter) warrant further investigation.

You need a precision scale with 0.001 g resolution. Digital callipers with 0.01 mm resolution handle diameter and thickness. You can also use the free Dimension Camera tool on the EON Bullion website for phone-based measurement against a known reference length.

Weight and dimension testing catches crude counterfeits made from the wrong metal or with incorrect sizing. However, it cannot detect tungsten-core fakes, which are manufactured to exact specifications. This is why dimensional checks should always be combined with conductivity testing.

Gold is diamagnetic — it is very weakly repelled by a magnetic field. A strong neodymium magnet should slide slowly down a tilted gold coin or bar, neither sticking nor falling freely. This sliding effect is caused by eddy currents induced in the gold, which create a gentle braking force.

The magnet slide test can detect ferromagnetic counterfeits (those containing iron, nickel, or steel) because these metals cause the magnet to stick. However, tungsten is non-magnetic, so a tungsten-core fake passes the magnet test just as easily as genuine gold.

While the magnet slide test is simple and inexpensive, it provides limited information. It should be considered a preliminary screening tool rather than a definitive authentication method.

When struck on its edge, a genuine gold coin rings with a clear, sustained tone. The frequency of this ring is determined by the coin’s composition, dimensions, and internal structure. Each coin type has a characteristic frequency that can be measured using a microphone and frequency analysis software.

Each coin type has a predicted ring frequency: the site’s database predicts about 4,830 Hz for a 1 oz Maple Leaf and 1,941 Hz for a 1 oz Philharmonic. A tungsten-core fake should ring at a different frequency; a tungsten-cored coin is predicted to ring higher, not lower, because tungsten is much stiffer than gold. Most reference values are predicted from plate theory, not measured.

Acoustic testing is non-destructive and can be performed with a smartphone. The EON Ping Tester uses the Web Audio API to capture and analyse the frequency spectrum in real time, comparing against predicted reference values.

Eddy current testing measures a property that a plated tungsten fake cannot match: the coin’s electrical conductivity through its thickness. It works by measuring the electrical conductivity of the entire sample volume, not just its surface. Pure gold has a conductivity of 44.0 MS/m, while tungsten is only 17.9 MS/m — a difference of more than 2:1 for fine-gold coins. For 22-carat coins (about 10 MS/m) the gap is smaller, and a layered fake can come close.

EON’s eddy current method uses a neodymium pendulum that swings past the coin. The eddy currents induced in the coin create a braking force that brings the pendulum to rest sooner. The reading depends on the coin’s conductivity, thickness and size, and is compared with a genuine coin of the same type. The free online simulator and optional Flat-Pack Pendulum Kit both use this principle.

Compare the reading with a measured genuine coin of the same type on the same unit; the repeatability there is about ±5%. Without a reference coin, the formula is only good to about ±21%. In the model, a gold-plated tungsten fake of a fine-gold coin falls well outside either band, because plating is too thin to mask the core. These fake results are modelled, not measured.

No single test should be relied upon in isolation. The most robust authentication combines multiple independent methods, each catching a different category of counterfeit. The EON multi-factor protocol uses four tests: dimensional verification, weight check, acoustic ping, and eddy current analysis.

The EON Multi-Factor Score combines the completed tests into a 0–100 consistency score: how closely your measurements match the published specifications. It is not a probability that the coin is genuine.

For private sales, where buyer and seller may be in different countries, shared measurements from several independent tests give both sides something concrete to check.