High-Grade Gold Discoveries: Latest Intercepts Compared

A high-grade gold discovery is a drill intersection that sits far above what mines of the same style average. Metals Focus put the average underground gold mine at 4.15 grams per tonne (g/t) and the average open-pit mine at 1.29 g/t in 2021 (Kitco, reporting Metals Focus, December 2022). Against that yardstick, the company-reported intercepts below, which include 10.32 g/t over 6.00 m, 32.4 g/t over 12.2 m and 61.5 g/t over 15.0 m, are high grade. A single hole is not a mine, though, and this page shows how to read the number, what the newest results are, and what to check before you act on one.

Grade is the amount of gold in a tonne of rock, in g/t. One g/t equals one gram of gold in a tonne of ore. The page is built around three questions: what counts as high grade, which discoveries have been announced lately, and how an explorer or investor can tell a strong result from a flattering one. All drill results below are reported by the companies, unaudited, and dated. Not investment advice.

What counts as high grade gold in g/t

No rule sets a single threshold. Neither JORC nor NI 43-101, the two reporting codes most explorers use, defines “high grade” (see our guide to reading a resource table). The useful test is relative: compare the number with the average for the same mining method, because underground mines can afford to treat much richer ore than open pits can.

The table shows where published averages and company figures sit. The last column turns each grade into the value of the gold in one tonne of rock at the Kitco spot bid of US$4,046.40 per troy ounce at 7:28 a.m. EDT on 28 October 2026 (Kitco). That is US$130.09 per gram, so each 1 g/t is worth about US$130 per tonne before recovery losses, royalties, tax and the cost of mining and milling.

Two points stand out. First, Bellevue’s reserve grade (4.5 g/t) is lower than its resource grade (8.6 g/t) in the same statement (Bellevue Gold, 21 September 2026). A reserve is the part of a resource shown to be economic after allowing for mining dilution and cut-offs, so its grade is usually lower. Second, a drill intercept is not a mine average. Mines blend rich and poor ore, so a mine’s average grade usually sits well below its best drill intercepts.

Older rankings show how far the top end can run. A 2015 analysis of 2014 reserve data put Fire Creek in Nevada first among underground mines at 44.1 g/t and Macassa in Ontario second at 22.2 g/t (MINING.COM, July 2015). Those figures are a decade old and ownership and reserves have changed since, so use them only as a sense of scale. Our page on the largest gold mines by reserves covers size rather than grade.

Chart

Published average grades versus one 2026 drill intercept

Grams of gold per tonne; averages for 2021 (Metals Focus) and the September 2026 Bellevue statement; the Goliath bar is a single intercept, not an average

Recent high-grade gold discoveries and their intercepts

These are the newest results we checked against the companies’ own releases, as of 28 October 2026. “Discovery” here follows the company’s wording: a zone found by drilling, not a mine. The grades are in g/t gold (Au) unless stated, and each interval is a length along the drill core.

Surebet and Golden Gate, Golden Triangle, British Columbia (Goliath Resources)

Hole GD-26-424 assayed 10.32 g/t Au over 6.00 m, including 15.61 g/t over 3.96 m and 33.80 g/t over 1.82 m, in quartz-sulphide veins with visible gold. The company says the Golden Gate Zone has been extended by 860 m this year (320 m north-east and 540 m south-west) and is still open (release of 11 August 2026). Another hole, GD-25-412, cut 4.41 g/t gold equivalent over 7.79 m in the Surebet Zone, including silver at 61.14 g/t, which is why the release uses the term AuEq, gold equivalent.

Sunday Creek, Victoria, Australia (Southern Cross Gold)

Hole SDDSC188 returned 12.2 m at 32.4 g/t Au from 447.0 m, with a single 0.2 m sample at 1,050 g/t, in the Golden Dyke area (January 2026 release). The company calls it the best gold intersection drilled at Golden Dyke so far and says the hole linked Golden Dyke with the nearby Rising Sun zone. This is a gold-antimony system, so the headline figure is quoted both as gold and as gold equivalent.

Duquesne West, Abitibi, Quebec (Emperor Metals)

Emperor Metals holds the property under option from Duparquet Assets, which is 50% owned by Globex. Hole DQ26-20 returned 61.5 g/t Au over 15.0 m of core with visible gold. The same release shows that two adjoining samples (557.5 g/t over 1.4 m and 100.9 g/t over 1.0 m) make up almost all of it, inside a wider 35.8 m zone that averages 26.09 g/t (Globex release, 16 June 2026). The next section uses this hole to show why the average hides the shape.

Abore, Asanko Gold Mine, Ghana (Galiano Gold)

Step-out drilling at Abore reported 15 m at 14.2 g/t Au and 4 m at 30.4 g/t Au from 500 m, plus 24 m at 4.7 g/t Au. The company says the results show mineralisation up to 200 m below earlier holes (Galiano Gold, 29 January 2026). Abore is a useful case because it shows the next step: Galiano’s full-year report refers to an underground Mineral Resource at Abore, with 2026 drilling aimed at supporting a potential maiden underground reserve (Galiano Gold, full-year 2025 results).

Bellevue, Western Australia (Bellevue Gold): infill, not a new find

Bellevue is a producing mine, so its results are infill drilling rather than a discovery. They still show what a high-grade system looks like: Deacon North holes of 9.3 m at 62.1 g/t, 4.6 m at 50.3 g/t and 8.9 m at 31.0 g/t, and a 3.4 m intersection at 18.45 g/t at Westralia that the company is following up with a downhole electromagnetic survey (Bellevue Gold, 21 September 2026).

Hopes Hill and Greenmount, Southern Cross belt, Western Australia (Golden Horse Minerals)

At Greenmount, reverse-circulation hole 26GMRC006 gave 8 m at 16.3 g/t Au from 209 m (reported earlier in 2026 and restated in the 19 August release). At Hopes Hill, diamond hole 26HHDD031 gave 16.6 m at 3.4 g/t Au from 392.4 m, including 5.0 m at 9.9 g/t, beneath the historic Hopes Hill pit (ASX announcement, 19 August 2026). The Hopes Hill diamond hole is lean in grade but wide, which matters for tonnage. The same release also reports 10 m at 11.3 g/t Au from 116 m (26HHRC147), including 2 m at 52.3 g/t, so Hopes Hill is not a low-grade project.

The chart plots each reported intercept by width and grade. Intercepts toward the top are the richest; those toward the right are the widest. The best targets are both, and they are rare.

Chart

Recent intercepts by width and grade

Selected 2026 intercepts, one per project or zone (best by gram-metres in the cited release, except Hopes Hill and Greenmount, where other holes are listed in the text); core lengths, not true widths; company-reported and unaudited

Gram-metres: comparing a short rich intercept with a long one

A gram-metre is grade multiplied by width: 10 g/t over 6 m is 60 gram-metres, the same as 5 g/t over 12 m. It is a quick way to compare holes, because it rewards both richness and thickness. On the reported figures, the Duquesne West hole is about 922 gram-metres (61.5 g/t x 15.0 m), Bellevue’s best Deacon North hole about 578, Sunday Creek about 395, Abore about 213, Goliath about 62 and Hopes Hill’s wide diamond hole about 56 (its RC hole 26HHRC147, 10 m at 11.3 g/t, is about 113).

The Duquesne West assay table shows why the arithmetic needs care. Adding up the table gives about 906 gram-metres, close to the headline average. Of that, 781 comes from the 1.4 m sample at 557.5 g/t, 101 from the 1.0 m sample at 100.9 g/t, and only about 25 from the other 12.6 m. Roughly 97% of the metal sits in 2.4 m of core. The release itself notes that the high-grade interval is surrounded by lower-grade material.

That does not make the result less real. It means the 15 m average describes a short, very rich vein inside a halo, not 15 m of uniformly rich rock. Mine planning, mining method and dilution all depend on that shape, so ask for the interval-by-interval table, not just the headline.

Chart

Two samples carry the Duquesne West intercept

Gram-metres (g/t x metres) by sample, hole DQ26-20; calculated from the assay table in the release

Core length, true width and the nugget effect

Drill intercepts are measured along the core, and the hole rarely cuts the vein at a right angle. True width is the thickness of the vein itself. Southern Cross Gold states that the true thickness of its Sunday Creek intervals is about 55% to 65% of the sampled thickness (January 2026 release). On that basis the 12.2 m intercept is roughly 6.7 to 7.9 m thick. At Duquesne West, the company says true widths are 70% to 90% of the reported widths, so 15.0 m of core is about 10.5 to 13.5 m of rock. Always find the true width statement; if there is none, treat the width as an upper limit.

Gold is also unevenly spread. Coarse gold grains can sit in a few samples, a pattern called the nugget effect. A few “nuggety” high assays can inflate the estimated gold in a deposit, which resource geologists manage by top-cutting, reducing values above a chosen limit (Datamine documentation). Visible gold in core, as reported by Goliath and at Duquesne West, is a good sign, but only the laboratory assay tells you how much gold the sample holds.

Check how the company composited the intervals, too. Southern Cross Gold says it reports lower grades using a 1.0 g/t gold-equivalent cut-off over a maximum of 2 m, and higher grades using a 5.0 g/t cut-off over a maximum of 1 m. Different rules give different headlines from the same core, so the rules matter when you compare one company with another.

Gold price, falling average grades and the cost of a mine

Higher prices make poorer rock worth mining, so average grades fall as prices rise. Metals Focus found that the average head grade at primary gold operations peaked at 1.46 g/t in 2017 and had fallen 8% to 1.35 g/t by 2021, partly because the rising price makes low-grade material economic (Kitco, December 2022). The USGS puts the average gold price at US$1,801 per ounce in 2021 and US$3,300 in 2025 (estimate), an increase of about 83% (USGS Mineral Commodity Summaries 2026).

Costs have risen too. The World Gold Council reports that the global average all-in sustaining cost (AISC) rose 5% on the quarter and 16% on the year to US$1,785 per ounce in the first quarter of 2026 (World Gold Council, August 2026). Set against the spot bid on 28 October 2026 the gap is wide, though the two figures are from different dates and an AISC average does not describe any single project.

The calculator turns a reported intercept into value per tonne, true width and break-even grade. It uses your inputs, including the price and cost, which are examples only.

Technology is changing how teams drill. Bellevue reports 193,000 m of underground grade-control drilling in the year to March 2026, and Galiano completed an 11,000 m Abore program in six weeks. Downhole electromagnetic surveys, such as the one at Bellevue’s Westralia target, help drill teams decide where to put the next hole. Better targeting raises the odds of a discovery, but it does not change the grade of the rock.

Chart

Average gold price, 2021 to 2025

US dollars per troy ounce; 2025 is a USGS estimate

From drill hole to mine: what has to happen after a discovery

A discovery hole is the first step. The usual path is more drilling to test continuity, a maiden resource estimate, technical studies, permits and finance, and then construction. At Abore, the 2025 drilling was timed to feed a maiden underground resource that Galiano expected in February 2026 (Galiano Gold, 29 January 2026), and 2026 infill drilling is meant to support a possible maiden underground reserve.

Infographic

From discovery hole to mine in 6 steps

1. 1Discovery holeA drill hole cuts high-grade gold and the lab assays it.
2. 2Step-out drillingMore holes test the zone along strike and at depth.
3. 3Maiden resourceA qualified person estimates tonnes and grade under a reporting code.
4. 4Technical studiesMetallurgy, mine design and costs show if the gold pays.
5. 5Permits and financeLicences, community agreements and funding are secured.
6. 6Build and mineConstruction starts and the ore is mined and milled.

Discovery matters because the stock of known gold is finite. The USGS counts world gold reserves at 66,000 t against 2025 mine output of about 3,300 t, which is roughly 20 years of supply at that pace (our arithmetic on USGS figures). It estimates that the United States holds 33,000 t of gold resources, of which 18,000 t is undiscovered (USGS Mineral Commodity Summaries 2026). Our pages on countries with the most gold reserves and top gold producing countries show where that gold sits.

▶ A clay nugget travels from rock to ring: how gold is found, drilled, crushed, dissolved and refined. · 5:27

What investors and explorers should check

Use this list on any high-grade gold discovery announcement. Most of the answers are in the release itself, or in the table of assays attached to it.

1. Interval table. Is there an interval-by-interval assay table, or only the headline average? Look for the Duquesne West pattern, where two samples carry the result.
2. True width. Does the release state true width, and what share of core length is it? Treat core length as a maximum if it does not.
3. Composite and cut rules. What cut-off and maximum dilution were used to build the interval, and were high grades capped?
4. Gold or gold equivalent. If the headline is AuEq, check the metals and prices behind the conversion, as with silver at Goliath or antimony at Sunday Creek.
5. How many holes. One hole is a lead. A pattern of holes along strike and at depth is a discovery. Ask for the number of holes drilled and how many hit.
6. Lab and sampling. Which lab, which assay method (fire assay is standard for gold) and what quality control? Visible gold in core does not replace assays.
7. Resource category. Is the figure an exploration result, an inferred or indicated resource, or a reserve? Each is a different level of confidence, and a reserve grade is usually lower than the resource grade.
8. Depth and mining method. A 5 g/t result at 600 m depth is an underground story with underground costs.
9. Metallurgy. What share of the gold can be recovered, and is the gold coarse and easy to recover, or locked inside sulphide minerals?
10. Land, permits and funding. Who owns the ground, what permits are needed, and how will the next stage be paid for? Our gold prospecting guide covers the deposit types that these results usually come from.

A Canadian release should name a qualified person, and an Australian one a competent person, who takes responsibility for the technical content. The company still chooses what goes in the headline, so read beyond it.

Where satellite screening fits

Satellites do not see gold or measure grade. They map the rock around it: iron-oxide staining, clay and mica alteration, silica and the faults that hot fluids used (our guide to epithermal gold deposits lists those clues). Screening an area this way can help rank ground before a claim is staked or a drill program is planned, and it is the cheapest step in the chain; drilling is what confirms grade. Our note on mining technology covers other tools.

▶ Why satellites cannot see gold, and how they map the rock around it. · 4:25

Frequently asked questions

What is considered high-grade gold?

There is no official threshold. As a guide, Metals Focus gave 4.15 g/t as the 2021 average for underground mines and 1.29 g/t for open pits, so as a rule of thumb, a grade several times those averages is called high grade. Bellevue’s 2026 reserve grade of 4.5 g/t shows where a producing high-grade underground mine can sit.

What is the latest gold discovery?

Among the results on this page the newest are Bellevue’s Deacon North infill holes (21 September 2026), Golden Horse’s Hopes Hill hole (19 August 2026) and Goliath’s Golden Gate hole (11 August 2026). Goliath has since issued releases on 10, 16, 18 and 22 September and 7 October 2026, so check its news page. Releases arrive weekly, so check the company’s news page and the exchange filing for anything newer.

Is 5 g/t gold good?

The 2021 open-pit average was 1.29 g/t and the underground average 4.15 g/t, so 5 g/t is about four times the open-pit average and slightly above the underground one. As a rule of thumb, that is rich for an open pit and good but not unusual underground. What matters is the width, the depth and the cost to mine it.

How do I read “10 g/t over 6 m”?

It means the sampled 6 m of core averaged 10 grams of gold per tonne. Multiply the two numbers for 60 gram-metres. It is a core length unless the company states a true width.

What is the difference between a discovery and a resource?

A discovery is an exploration result, such as one or several drill holes. A resource is a formal estimate of tonnes and grade prepared under a code such as JORC or NI 43-101, and it needs enough drilling to show continuity. A reserve is the part of a resource that is economic to mine.

What does “high grading” mean in mining?

It has two meanings. Historically it meant miners stealing rich ore, for example in lunch pails, in the early 1900s (A Way with Words). In planning, it means mining the richest parts of an orebody first, which lifts early output but leaves a lower grade for later years.

Where are the highest-grade gold mines?

A 2015 ranking of 2014 reserve data led with Fire Creek in Nevada at 44.1 g/t, ahead of Macassa in Ontario at 22.2 g/t (MINING.COM). Check each owner’s latest reserve statement before relying on those numbers. See also gold mines around the world.

Does a high-grade intercept mean a mine will be built?

No. Most intercepts never reach a mine. A mine needs enough tonnes, recovery, permits, finance and a price above cost. Bellevue’s resource grade fell from 8.9 to 8.6 g/t between statements (and its reserve grade from 4.7 to 4.5 g/t), which shows that stated grades move from year to year.

Can satellites find high-grade gold?

Not by itself. Satellites map alteration minerals and structures that point to where gold may have formed. Only drilling and laboratory assays measure grade.

How much gold is in a tonne of 10 g/t ore?

Ten grams, which is about 0.32 troy ounces. At the Kitco bid of US$4,046.40 per ounce on 28 October 2026 that is about US$1,301 of contained gold, before recovery losses, royalties and costs.