How to Read the EIA Drilling Productivity Report

Every week, the Baker Hughes count tells the market how many rigs are drilling in the United States. Every Wednesday, the EIA Weekly Petroleum Status Report tells the market how much oil the country produced, in round modelled numbers. Between those two releases sits a monthly EIA publication that connects them: the Drilling Productivity Report, or DPR. The DPR takes the rig count for seven shale regions, estimates how much new production each rig is bringing on, subtracts the decline of every older well in the region, and publishes the result as an indicated change in oil and gas production. It is, in effect, the EIA answer to the question the rig count cannot answer on its own: so how many barrels does that rig count mean?

This guide walks through the report in the order a careful reader uses it. It is the third guide in our EIA reading series, after the Weekly Petroleum Status Report guide and the Weekly Natural Gas Storage Report guide. Those two reports count what is in tanks. This one estimates what is coming out of the ground next, and why the total keeps trying to fall. Every figure in the tables below comes from the EIA Drilling Productivity Report page at eia.gov, checked on October 18, 2026, or from PetroEyes checked-in DUC data file, which reproduces the same EIA DUC table. Where a worked example uses invented teaching numbers instead of report data, it is labelled a hypothetical, in those words.

1. What the DPR is, and what it is not

The EIA describes the method in one sentence: the report uses recent data on the total number of drilling rigs in operation, together with estimates of drilling productivity and estimated changes in production from existing wells, to provide estimated changes in oil and natural gas production for the key regions. Three inputs, one output. The rig count is observed. The productivity and decline figures are modelled from state and company production data that arrive with a lag, and the EIA applies smoothing techniques to most of the series because the underlying data are noisy. The DPR is therefore an estimate built on estimates, published because the alternative, waiting for final state production records, would leave readers months behind the shale fields it describes.

Two limits follow from that design, and the EIA prints both on the report page itself. First, the headline productivity metric, new-well production per rig, can become unstable during periods of rapid increases or decreases in rigs and completions. Second, the legacy decline metric can become unstable when large volumes of production are curtailed or shut in, as happened during winter freeze-offs, flooding, and the 2020 demand collapse. A reader who knows those two warnings in advance will not be surprised when a storm month or a price crash produces a DPR print that later revisions quietly reshape.

The DPR also does not distinguish between oil-directed and gas-directed rigs. The EIA gives the reason directly: once a well is completed it may produce both oil and gas, and more than half of the wells in these regions produce both. Each regional table therefore carries an oil side and a gas side for the same rig fleet. That is why the Haynesville, a gas region, can show a new-well oil figure of just 14 barrels per day per rig next to a gas figure above 13,000 thousand cubic feet per day per rig in the table in section 4, and why Appalachia shows the mirror image in reverse. The rig is the unit of effort. The rock decides what comes out of it.

2. The seven regions

The DPR covers seven regions: Anadarko, Appalachia, Bakken, Eagle Ford, Haynesville, Niobrara, and Permian. The EIA chose them because they drove the shale era: by the agency account, these regions accounted for 95 percent of domestic oil production growth and all of the domestic natural gas production growth during 2011 to 2013, the years when US shale output was reshaping world markets. The set has been stable since the Anadarko and Appalachia regions were added to the original five, which makes the DPR one of the few long, consistent regional series in US shale data.

The regions are not equal, and the production table in section 4 shows the scale gap in the EIA own numbers. In June 2024, the most recent production month posted on the EIA DPR page when this guide was checked, the Permian produced 6,187 thousand barrels of oil per day out of a seven-region total of 9,853 thousand barrels per day, which is about 63 percent of the total from one region. Appalachia produced just 135 thousand barrels of oil per day but 35,791 million cubic feet of gas per day, the largest gas figure in the table. Reading the DPR nationally, without the regional split, mostly means reading the Permian for oil and Appalachia plus the Permian for gas. The other five regions still matter, because their declines and recoveries are where turning points show up first, but their weight in the total is what it is, and the table makes it visible.

Region definitions also decide what the report can and cannot see. Production outside the seven regions, including the Gulf of Mexico, conventional onshore fields, and smaller shale plays, does not appear in the DPR at all. The seven-region oil total of 9,853 thousand barrels per day in June 2024 is therefore not US production; it is the shale-region subset of it. Readers who need the national figure should use the production block of the Weekly Petroleum Status Report and the benchmarks on our WTI price page, and treat the DPR as what it is: a regional supply engine room, not a national balance sheet.

3. The three tables, in the order to read them

Each DPR release is built from three tables. Commentary tends to quote the first and ignore the other two. The reading order that works is production first, productivity second, and DUC wells third, because production is the conclusion, productivity is the main assumption behind it, and the DUC count is the forward check on whether the assumption can hold.

Table 1: production by region, and the indicated change

For each region the report prints total oil production and total gas production for the current month and the next month, with the change between them. That change, the indicated monthly change, is the single most useful number in the report. It is not measured directly. It is computed, as section 5 shows, by adding the estimated production from new wells to the estimated change from legacy wells. When you quote a DPR production figure, you are quoting the output of that addition, one month of which is always a projection past the latest drilling data.

Table 2: new-well production per rig

For each region the report prints new-well oil production per rig, in barrels per day, and new-well gas production per rig, in thousand cubic feet per day, for two consecutive months. This is the productivity trend line: it captures longer laterals, better targeting, higher-intensity completions, and also the move into weaker rock, all in one ratio. Section 6 explains why the ratio is built the way it is, including the two-month lag that surprises most first readers.

Table 3: drilled but uncompleted wells

The third table counts DUC wells by region: wells that have been drilled but not yet completed, and so are not yet producing. The DUC count is the inventory between drilling and production. Our dedicated explainer on drilled but uncompleted wells covers the well lifecycle in full. In the DPR, the DUC table earns its place because it is the only forward-looking stock in the report: a large DUC inventory lets completions, and therefore new production, continue for a time even if drilling slows, while a thin inventory ties future production tightly to the current rig count.

4. The report tables, laid out in full

The two tables below reproduce the most recent figures posted on the EIA Drilling Productivity Report page at eia.gov when this guide was checked on October 18, 2026: new-well production per rig and production by region for May and June 2024, shown here as the June 2024 level and the change from May. They are the original working asset for this guide. Read them together: the first table is the productivity assumption, the second is what the assumption produced.

Three readings fall out of the pair. First, the Permian did almost all of the oil work: its indicated oil change of +18 thousand barrels per day is larger than the seven-region net of +17, which means the other six regions, combined, slightly declined. Second, productivity and production are different stories: the rig-weighted average for new-well gas production per rig fell by 133 thousand cubic feet per day even as Permian gas production rose by 143 million cubic feet per day, because production depends on rig numbers and legacy decline as well as on per-rig productivity. Third, the changes are small against the levels. A +17 thousand barrel per day month on a 9,853 thousand barrel per day base is a plateau month, and the DPR is at its most useful exactly there, because it shows which side of the ledger, new wells or legacy decline, is winning by how little.

5. The core equation: new wells against legacy decline

Every number in the production table comes from one addition:

Indicated production change = production from new wells + legacy production change

Shale wells decline fast. A typical shale oil well produces its highest rate in its first months and then falls steeply, which means a region standing still is a region whose existing wells are collectively losing output every single month. The EIA calls that loss the legacy production change, and in shale regions it is negative in normal times. New wells must first replace the legacy loss before they add a single net barrel. Analysts call the effort required just to hold output flat the treadmill, and the DPR is the report that measures the speed of the treadmill, region by region, month by month.

Here is the arithmetic in a labelled hypothetical, using invented teaching numbers at a realistic Permian scale, not report data. Suppose a region runs 250 rigs, the applicable new-well oil production per rig is 1,400 barrels per day, and the legacy decline for the month is 332 thousand barrels per day. New-well production is then 250 multiplied by 1,400, which is 350 thousand barrels per day of gross additions. The indicated change is 350 minus 332, which is +18 thousand barrels per day. Note what the hypothetical shows: a gross effort of 350 thousand barrels per day, from 250 rigs working a full month, nets to 18. About 95 percent of the new production replaced decline. That ratio, not the rig count and not the net change alone, is the honest description of a mature shale region, and it is why small shifts in either term flip the sign of the total. Drop the rig count by 10 in the same hypothetical and gross additions fall to 336 thousand barrels per day, and the indicated change becomes +4. Drop productivity by 100 barrels per rig per day instead, and the same flip arrives by a different road.

The gas side works the same way with bigger units and, in Appalachia and the Haynesville, much larger per-rig figures. It also explains a pattern that confuses new readers of our rig counts page: production rising for months after the rig count has peaked. Completions of already drilled wells keep arriving, per-rig productivity keeps improving, and the new-well term can exceed legacy decline for a long time on momentum built up when rig numbers were higher. The reverse also holds. A rising rig count does not lift production until the drilled wells are completed, which is the lag the next section formalises.

6. New-well production per rig: read the definition before the number

The EIA states the construction of its headline metric in one paragraph, and the paragraph matters more than any single print. The metric uses a fixed ratio of estimated total production from new wells divided by the region monthly rig count, lagged by two months. Two consequences follow. The lag exists because a rig that spuds a well today delivers first production weeks to months later, after drilling, completion, and connection. Applying a two-month-old rig count to this month new production is the report way of matching cause to effect. And the figure is a regional average across every well a rig fleet brought on, not the initial rate of any individual well, and not, as the EIA underlines, production per newly completed well.

Read as a trend, the metric is the cleanest public record of shale efficiency. In the June 2024 table in section 4, Bakken new-well oil production per rig stands at 1,747 barrels per day and Permian at 1,400, against a rig-weighted average of 1,222. A decade earlier, in the February 2014 DPR table archived on the same EIA site, the Permian figure was 90 barrels per day and the Bakken figure was 479. Longer horizontal laterals, multi-well pads, and more intense completions explain most of that rise, and our Baker Hughes rig count explainer walks through why rig counts stopped being a stand-alone production forecast as this ratio climbed. Two rigs in 2024 were not two rigs in 2014, and the DPR is where that difference is quantified.

Read as a monthly signal, the same metric needs care. Because the denominator is a lagged rig count and the numerator is modelled new production, a month in which operators high-grade their best remaining locations can lift the ratio without any technology change, and a month in which completions bunch can do the same. The EIA warning bears repeating in this section: the metric can become unstable during periods of rapid decreases or increases in the number of active rigs and well completions. Use the ratio for the direction of efficiency over quarters. Use the indicated production change, which folds the ratio back into rig numbers and decline, for the monthly supply call.

7. DUC wells: the inventory identity and the April 2024 table

The DUC table obeys an identity simple enough to check by hand:

Change in DUC wells = wells drilled in the month - wells completed in the month

The table below is computed at build time from PetroEyes checked-in DUC data file (src/data/json/duc_data.json), which reproduces the EIA DPR DUC table for April 2024, the most recent DUC month posted on the EIA DPR page when this guide was checked. Every row satisfies the identity, and the regional rows sum to the DPR total, subject to the file region definitions.

In April 2024 the seven regions drilled 864 wells and completed 858, so the DUC total rose by 6 to 4,510 wells. The Permian held 893 of them, the largest single regional stock, and the Niobrara posted the largest monthly draw, completing 94 wells against 78 drilled. A month this close to balance, with drilled and completed within single digits of each other nationally, describes a system in steady state: operators are completing about what they drill, neither building a completion backlog nor spending one down. The DUC level then matters more than the DUC change, because the level is the buffer available if prices or service costs move and operators want to change completions faster than they can change drilling.

One filing caution belongs here, and it is visible in our own data file. The checked-in file also carries a headline total that sums regional and multi-region series together; the figure used in this guide is the DPR regions row of 4,510 wells, which matches the total printed in the EIA table itself for April 2024. When two totals appear in a dataset, the one that matches the source table, and that satisfies the drilled minus completed identity against its own prior month, is the one to quote. That check takes a minute and prevents the most common DUC error in secondary coverage, which is quoting a summed series as if it were the report total.

8. Three clocks: the rig count, the DPR, and the weekly report

US shale data runs on three clocks, and most reading errors come from expecting one clock to do another clock job. The Baker Hughes rig count is the weekly clock: it counts active rigs every Friday, it is observed rather than modelled, and it leads production by the drilling and completion lag. The DPR is the monthly clock: it converts rigs into estimated production changes by region, with productivity and decline modelled in. The Weekly Petroleum Status Report is the weekly production clock: its domestic production figure is a rounded modelled estimate that the monthly surveys later benchmark, as our guide to that report explains in section 3 of its reading order.

Used together, the three clocks form a chain of confirmation. A rig count turn in, say, the Permian should appear in the DPR as a change in new-well production roughly two months later, and in the weekly production estimate as a drift in the same direction over the following quarter, with the monthly Petroleum Supply Monthly as the final arbiter long after. When the chain breaks, when rigs rise and the DPR shows no new-well response, the break itself is the story: completions are being deferred, the DUC count should be rising to prove it, and section 7 identity is where you check. Our Permian Basin production outlook applies exactly this chain to the largest region, and our EIA inventory terminal archives the weekly end of it.

The practical rule is to match the claim to the clock. A claim about drilling intent this month belongs to the rig count. A claim about whether shale supply grows next month belongs to the DPR indicated change. A claim about what the United States produced last week belongs, with its error bars acknowledged, to the weekly report. Swapping those assignments, quoting a rig count as a production forecast or a DPR projection as a measured outcome, is how confident and wrong shale commentary gets written.

9. Five common misreads

Misread 1: More rigs means more production next month. Rigs drill wells; completed wells produce. With a two-month productivity lag in the DPR itself and a DUC stock in between, a rig increase is a production story for a later quarter, and a rig decrease can coexist with rising production for months while completions work through the backlog.

Misread 2: New-well production per rig is the output of one new well. It is a ratio of total estimated new production to a lagged rig count. The EIA says so in terms. Quoting it as a per-well initial production rate overstates some regions and understates others, and breaks completely in gas regions where the oil ratio is near zero by design.

Misread 3: A positive indicated change means the region is healthy. The sign hides the treadmill ratio. A region adding 350 thousand barrels per day of new production to net +18 is working far harder for its growth than a region adding 60 to net the same +18. Section 5 hypothetical shows the arithmetic. Always ask how much gross addition the net required.

Misread 4: The DUC count is spare production waiting to flood the market. DUC wells still need completion crews, frac sand, water, and pipeline connections, and some drilled wells are never completed because the rock disappointed. The DUC stock is a buffer measured in months of completions, not a tap. Divide the regional DUC count by its monthly completions to size it honestly: at the April 2024 completion pace in the table above, the Permian stock of 893 wells equals about two months of Permian completions.

Misread 5: The DPR total is US production. The seven regions produced 9,853 thousand barrels per day in June 2024 in the EIA table. US production in the same period, as tracked in the weekly report, was materially higher, because the DPR excludes offshore and conventional supply by design. Keep the subset label attached to the number whenever you repeat it.

10. Your monthly checklist

Work down the rows in order when a new DPR lands. Each row names where to look and what the check settles.

11. Frequently asked questions

What is the EIA Drilling Productivity Report?

It is the EIA monthly report that estimates how oil and gas production is changing in seven US shale regions. For each region it combines the rig count with estimates of drilling productivity and the decline of existing wells to produce an indicated change in production, and it counts drilled but uncompleted wells. It is the monthly companion to the weekly rig count tracked on our rig counts page and to the Wednesday petroleum report covered in our Weekly Petroleum Status Report guide.

What does new-well production per rig actually measure?

It is an estimate of the total first-month production from the new wells one rig brings on, divided using the region rig count lagged by two months. The EIA states plainly that it does not represent production per newly completed well, and that the metric can become unstable when rig counts or completions change rapidly.

What is legacy production change?

It is the estimated monthly change in output from wells that were already producing before the new wells arrived. In shale regions it is almost always negative, because shale wells decline steeply after their first months. Regional production grows only in months when new-well additions exceed the legacy decline, as the worked example in section 5 shows.

How is the DUC count calculated each month?

By an inventory identity: wells drilled in the month minus wells completed in the month equals the change in the DUC count. In April 2024, the most recent month in the EIA table reproduced in section 7, the seven regions drilled 864 wells and completed 858, so the DUC total rose by 6 to 4,510 wells.

Why does the DPR use a two-month rig lag?

Because a rig drilling today does not produce today. Weeks pass while the well is drilled, and more time passes before completion and first production. The DPR therefore applies the rig count from two months earlier when it estimates the new production arriving in the current month.

What is the difference between the DPR and the Baker Hughes rig count?

Baker Hughes counts rigs each week; it says nothing directly about barrels. The DPR converts rigs into estimated barrels by adding a productivity estimate and a legacy decline estimate for each region. The rig count is the faster signal, the DPR is the translation of that signal into production terms, and the two can point in different directions in the same month. Our rig count explainer covers the weekly count itself.

12. Related reading

• How to Read the EIA Weekly Petroleum Status Report - the Wednesday companion guide: stocks, refinery runs, and days of supply.
• Baker Hughes Rig Count, Explained - the weekly input the DPR converts into production terms.
• Drilled but Uncompleted Wells, Explained - the well lifecycle behind the DUC table.
• Rig Counts - the current weekly count by country and basin.
• Permian Basin Production Outlook 2026 - the DPR chain applied to the region that carries the total.