Oil and Gas Rig Counts

The current count

Baker Hughes North America Rig Count, released October 23, 2026, as reported by the trade press on the day of release — see the provenance note, which records that BOTH of the publisher's own surfaces, the report file listing and the summary table, still displayed 10/16/26 when this was read, week ended October 23, 2026. Shares are computed on this page from the counts shown.

UNCHANGED at 603 for a second consecutive week, and for the second consecutive week the headline is the least informative figure in the row. Oil rigs rose 1 to 460, gas rigs fell 1 to 132, and miscellaneous rigs held at 11, so a zero on the week is two component moves cancelling. TWO CONSECUTIVE ZEROES ARE NOT THE SAME WEEK TWICE, and that is the whole reading. Last week's zero was three moves cancelling in the other direction — oil down 3, gas up 1, miscellaneous up 2 — so across the fortnight the oil-directed count has clawed back one of the three rigs it gave up and the gas count has given back one of the two it gained. A reader taking the headline alone would conclude the fleet has been static since October 9. It has not been static; it has been offsetting. The miscellaneous line holds at 11, which retires the question last week's entry left open rather than answering it: a 2-rig rise in the one component that had not moved at all could have been a one-week artifact, and a second week at 11 says it was not. That is recorded and not interpreted; it is two rigs on a line of eleven. Five arithmetic checks reconcile, each on reported components rather than derived ones: 460 plus 132 plus 11 is 603; the weekly component changes of plus 1 oil, minus 1 gas and no change in miscellaneous sum to zero, which is the reported weekly change, and 603 plus 0 is 603; the year-over-year component changes of plus 40 oil, plus 11 gas and plus 2 miscellaneous sum to the reported plus 53; the implied year-ago split of 420 oil, 121 gas and 9 miscellaneous sums to 550, which is 603 less 53; and all three of the release's component changes tie to the figures this row carried last week — 459 plus 1 is 460, 133 less 1 is 132, and 11 unchanged is 11. That fifth check is the strongest of the five. The publisher derives its weekly change column from its own prior-week file, so agreement across all three components is the publisher's prior-week file agreeing, component by component, with a figure this ledger entered from the wire seven days before the publisher's own table caught up to it. THE YEAR-AGO COMPARATOR MOVED AND IT IS NOT A DISCREPANCY. The publisher's line for the week ended October 16 reads plus 55 against 17 Oct 2025, which implies a year-ago 548; this week reads plus 53 against 24 Oct 2025, which implies 550. Those are comparators for two different weeks of 2025 — the comparator advances alongside the count — so the 2-rig rise belongs to the 2025 series and is not a disagreement between two Institute entries. Basin detail. The Permian holds at 272, up 22 against a year-ago 250. The Eagle Ford rose 2 to 51, up 8 on the year, which is its first move after three consecutive weeks at 49. Canada rose 3 to 220 on the same release, which ties to the 217 this ledger carried for the week ended October 16: 217 plus 3 is 220. North America is therefore 823 — 603 plus 220 — against a year-ago 749, being 550 plus 199, so the continent is up 3 on the week and up 74 on the year. Canada and the North America aggregate are recorded here as context for a North America release and are not fields this row prints. Context the row does not print as a field either: EIA's weekly estimate of U.S. crude production for the week ending October 16, 2026 is 13.269 mb/d. A YEAR-OVER-YEAR CHANGE ON THAT FIGURE IS NOT ASSERTED HERE. Two different year-ago readings were carried in the material read on this sweep and they imply different deltas, so the comparator is left unstated rather than printed on the stronger of two readings — a derived figure whose own inputs disagree is worth less than the absence of one. Primary Vision's frac spread count fell 4 to 199, its first fall in six weeks and still up 24 on the year; a rig count flat on the week and a completion count down on the week are measurements of two different stages of the same programme and do not have to agree. Trend: 588 on September 4, then 591, 595, 599, 598, 603, 603 and 603 — up 15 across seven weekly steps and flat across the last two. SUPERSEDED AND LABELLED rather than silently replaced, per the two-clock rule: the figure this row carried until today was 603 total — 459 oil, 133 gas, 11 miscellaneous — on the week ended October 16, 2026, unchanged on that week and up 55 against a year-ago 548. That reading was correct on its own clock, has since been confirmed on its total, its weekly change and its year-ago comparator against the publisher's own summary table, and is retained in the history table below.

Recent weeks

This table holds only the weeks the Institute has independently confirmed against a named report. It is deliberately short rather than deliberately long: a rig-count history assembled from secondary summaries acquires errors at exactly the weeks a reader would most want to trust it. Baker Hughes publishes the complete series, and a reader needing the full history should take it from the source. What this page adds is not the series. It is the reading of it below.

The price signal arrived and the rigs did not

Brent reached $126 on April 30, 2026, a four-year high. West Texas Intermediate posted a 35% weekly gain in early March, the largest since the contract began trading in 1983. Brent ended July at $96.80 after rising more than twenty-five dollars inside the month. By any account of how the American oil industry works that was written in the last fifteen years, that is a drilling boom.

The American rig count went from 603 in the week ended July 10, 2026 to 603 in the week ended September 4, 2026. That is 0 rigs across roughly eight weeks, and in the most recent two weeks the total did not move at all. Worldwide, the monthly average was 1,919 in September 2026, up 5.9% year over year. A forty per cent move in price has produced a 0.0 per cent move in the American count.

There are several explanations and they are not mutually exclusive. Shale operators spent the years after 2020 rebuilding balance sheets and returning capital, and boards that were punished for growth are slow to be rewarded for it. Service costs and rig availability constrain how fast a count can rise even when the decision is made. Efficiency means a rig today drills far more footage than a rig a decade ago, so a flat count is not flat activity. And the price is high because of a war, which makes it a price nobody underwrites a ten-year drilling programme against.

The Institute does not need to adjudicate between those explanations to state the consequence, which is the same under all of them: the rig count has stopped functioning as the market's automatic stabiliser. The reflex that shale drilling would cap any price spike rested on a response time of two to three quarters. In a disruption that has run since March, two to three quarters has already elapsed and the count has moved by 0.

Worldwide

The worldwide count is monthly rather than weekly and excludes several significant producing regions, most consequentially Russia and onshore China. It is a directional series for the rest of the world, not a global census, and the year-over-year change is the part of it worth reading.

The rest of the reference