We Counted Every 2026 Recall Where a Car Can Suddenly Lose Power While Driving

Key facts

• 33 recall campaigns filed in 2026 with an explicit "loss of drive/motive power / propulsion" consequence, from 566 total 2026 vehicle recall campaigns in the NHTSA flat file (strict filter; count as of 2026-09-12 and still growing).
• ~1,328,908 vehicles potentially affected across those 33 campaigns.
• Largest single campaigns: Nissan Rogue 26V080000 (323,917), Nissan Rogue 26V081000 (318,782), Jaguar Land Rover 26V248000 (170,169), Ford Mustang 26V547000 (148,663), Jeep Cherokee 26V290000 (61,711).
• Toyota C-HR EV — 26V570000 (Toyota 26TA17): 8,521 model-year 2026 battery-electric units; a BEV ECU software error can overcharge the traction battery and shut down the electric drive; report received 2026-09-02, owner letters 2026-10-18, remedy is a free ECU reprogram.
• Ford Mustang — 26V547000 (Ford 26C40): 148,663 model-year 2024-2026 Mustangs; engine-compartment wiring-harness ground eyelets can fracture, causing loss of drive power (plus headlights, washer, A/C, cooling fan) and failing FMVSS 108 and 104; report received 2026-08-25, remedy anticipated March 2027.
• Model years most represented (strict): MY2025 (14 campaigns), MY2026 (14), MY2024 (12), MY2023 (11), MY2022 (9).
• Excluded: the Ford Edge 2.0L EcoBoost coolant-intrusion item is an investigation (reported as defect petition DP26008), not a recall, and does not appear on any VIN check.

Which 2026 recalls involve a car losing power while driving?

Thirty-three 2026 recall campaigns describe a defect that can cut motive power while the vehicle is being driven, together covering about 1.33 million vehicles. These are manufacturer-acknowledged defects filed with NHTSA on Part 573 reports, not owner complaints. The mechanisms vary widely — seized engine bearings, broken throttle-body gears, fractured wiring grounds, failed DC-DC converters, cracked driveshafts, and overcharged EV batteries — but the consequence text is the same: an unexpected loss of drive power that raises crash risk.

The 12 largest by units affected:

The full 33-row dataset is published as a CSV (linked in Sources). These recalls have a close cousin in owner-reported data: our companion count of used cars with the most stalling and loss-of-power complaints shows the same failure described by drivers before it becomes a Part 573 report. The EV entries line up with our earlier VIN walkthrough of the used Toyota bZ4X problems to check before buying, which is campaign 26V393000 in the table above.

Which brand has the most 2026 power-loss recalls?

Ford is named in more 2026 loss-of-power recall campaigns than any other brand, and that result holds no matter how the filter is drawn. Under the strict filter (defect text explicitly says loss of drive/motive power), Ford is the sole named manufacturer on 12 of the 33 campaigns, and is named on 14 once the two it shares with Lincoln are counted; under a broader filter that also catches stall and engine-shutdown language, Ford appears in 17 of 90. Both brand-count definitions are stated in the Method section, because a reader re-running the filter will otherwise get a different number. No other make leads both. The newest model years dominate under both filters, which is expected: 2026 is a current recall year and its vehicles are still in production and warranty.

Because "power loss" is described in many ways, we ran the analysis twice. The strict count (33 campaigns / 1,328,908 units) is the defensible headline. A broader keyword set that adds stall, engine shutdown, sudden power loss and lose power returns 90 campaigns and roughly 3,626,604 units — but it also catches cases where a stall is a secondary consequence, so we treat it as an upper bound, not the number. The one thing that changes between filters is the runner-up: Nissan is #2 under strict, Mercedes-Benz under broad, so we assert no non-Ford superlative.

By the numbers

• 566 — distinct 2026 vehicle recall campaigns in the flat file as of 2026-09-12.
• 33 / 1,328,908 — campaigns / potential units under the strict "loss of drive/motive power" filter.
• 90 / ~3,626,604 — campaigns / units under the broader stall-and-shutdown filter (upper bound).
• 28 / 1,314,536 — strict campaigns / units after removing heavy trucks, buses, a motorcycle and an RV; the remaining 5 campaigns (~14,372 units) are commercial or two-wheel. Ford leads either way.
• 12 / 33 (sole make), 14 / 33 (named, incl. two shared with Lincoln) and 17 / 90 — Ford's share, strict and broad.

Based on 566 distinct 2026 vehicle recall campaigns (9,688 make/model/year rows) in the NHTSA FLAT_RCL_POST_2010 flat file, retrieved 2026-09-12; power-loss campaigns identified by case-insensitive keyword match on the Part 573 defect and consequence text and deduplicated by campaign number, counting each campaign's potentially-affected units once. Totals overstate real-world exposure because a single VIN can appear in more than one campaign, and 2026 is incomplete, so these counts will rise through year-end.

Is the Ford Edge power-loss issue a recall? (No — and it won't show on a VIN check)

The Ford Edge 2.0L EcoBoost coolant-intrusion issue is an open investigation, not a recall, so it is deliberately excluded from the 33-campaign count and will not appear on any VIN recall lookup. It is reported as NHTSA defect petition DP26008, covering roughly 499,603 model-year 2015-2018 Edge vehicles, with about 1,563 incidents of coolant entering the cylinders and causing a sudden loss of motive power (often with an "Engine Coolant Over Temperature" warning). The petition was submitted 2026-08-05 and is under evaluation.

This is the practical difference buyers keep hitting: a recall (Part 573) attaches to a VIN and shows up on a check; an investigation (PE / EA / defect petition) is NHTSA studying whether a recall is warranted and maps to a make/model/year population, not to your VIN. If NHTSA later upgrades the Edge petition into a recall, then affected Edge VINs would surface. Until then, watch NHTSA's investigations pages — a VIN check cannot track it. We break down that gap in detail in what "recall closed" vs. an open NHTSA investigation means on a VIN check.

What a VIN check can and can't tell you here

A VIN check confirms whether a specific used car is inside one of these open recalls; it does not confirm the repair was done, and it does not surface investigations. Enter the 17-digit VIN in NHTSA's free tool (the authoritative source) or a VIN-history report, and for the recall campaigns in this piece the check will show recall presence and count.

If you are shopping one of the affected models, run the VIN, then call a franchise dealer with the same VIN to confirm the specific recall is closed before you buy — the reason a presence flag is not the same as a fix is spelled out in what "recall completed" means when buying a used car. For screening open recall presence alongside a car's accident, salvage-auction, odometer, theft, ownership and past sales-listing history in one place, a Zilocar VIN check covers the history layer while you cede remedy confirmation to the dealer and investigation tracking to NHTSA. That open-flag-before-a-test-drive logic is the same one behind the Ford Bronco Sport do not drive recall check by VIN.

Method

Source file. NHTSA Office of Defects Investigation Recalls flat file FLAT_RCL_POST_2010.txt, extracted from FLAT_RCL_POST_2010.zip, retrieved and processed on 2026-09-12. The file is tab-delimited, one row per campaign / make / model / model-year combination, with no header row. Freshness was asserted before computing anything: the newest Part 573 report date in the file is 2026-09-09, three days before the run, and the archive's CRC verified intact. This check exists because NHTSA has re-published an ODI flat file under a newer timestamp with years of records missing; a stale-but-parseable file is the failure mode to guard against, not a missing one.

Universe. Keep RCLTYPECD = 'V' (vehicle recalls, excluding equipment, tyre and child-seat campaigns) and RCDATE beginning 2026. That leaves 566 distinct 2026 campaigns across 9,688 make/model/year rows.

The strict filter — the load-bearing definition. A campaign qualifies when its Part 573 defect text (DESC_DEFECT) or consequence text (CONEQUENCE_DEFECT) contains an explicit loss-of-propulsion phrase, matched case-insensitively: loss of drive, loss of motive, loss of propulsion, lose drive power, lose motive power. Nothing is matched on component codes or on our own reading of the mechanism — only on the manufacturer's own filed wording. That yields 33 campaigns / 1,328,908 potentially affected units.

Deduplication and counting. Campaigns are deduplicated on CAMPNO, and each campaign's POTAFF is counted once, not once per model-year row. Every figure here is a count of campaigns and of manufacturer-declared potentially-affected units — never of VINs, owners or confirmed failures.

Reproduction. The strict result was independently recomputed from a separately downloaded copy of the archive on the same day, on a copy NHTSA had re-published mid-run (245,578 rows against the 245,576 the first pull saw). The 33 campaigns and the 1,328,908 total reproduced identically on both copies, as did the model-year distribution (MY2026 14, MY2025 14, MY2024 12, MY2023 11, MY2022 9). Every campaign id and unit count in the frozen CSV was checked back against the flat file: 33 of 33 match, with no duplicate campaign ids.

Check 1 — filter sensitivity, the check most likely to move the number. "Power loss" is described many ways, so the analysis was run twice. A broader keyword set that also catches stall, engine shutdown, sudden power loss and lose power returns roughly 90 campaigns / ~3.63M units. That set deliberately catches campaigns where a stall is a secondary consequence of some other defect, so it is published as an upper bound and is not the headline. The strict number is the one to quote. What survives both filters is the direction: Ford leads under either, and the runner-up is not stable — Nissan under strict, Mercedes-Benz under broad — which is why no non-Ford superlative is asserted anywhere on this page.

Check 2 — vehicle class. The strict set includes vehicles a used-car buyer would never shop. Removing the five commercial, two-wheel and RV campaigns — Kenworth/Peterbilt 26V485000 (5,925), KTM 26V196000 (7,974), Storyteller Overland 26V274000 (357), Oshkosh 26V311000 (76) and Nova Bus 26V275000 (40), together 14,372 units — leaves 28 campaigns / 1,314,536 units. Ford leads either way. Both cuts are published so the reader can pick the universe.

Check 3 — how the brand count is defined, because a re-pull will otherwise disagree. Ford is the sole named manufacturer on 12 of the 33 strict campaigns. Counting the two campaigns it shares with Lincoln (26V091000, 26V122000), Ford is named on 14. Both numbers are correct under their own definition and both are far ahead of the next brands (Toyota and Nissan, 3 each), so the leadership finding does not depend on the choice. Anyone re-running the filter and counting any make string containing "FORD" will get 14.

Known limits. (1) These are manufacturer-filed defect reports, not confirmed failures — POTAFF is the population a manufacturer declares potentially affected, and the real defect rate within it is almost always far lower. (2) Units overstate real-world exposure: a single VIN can sit inside more than one campaign, and the 33 campaigns are not a count of 1.33 million distinct vehicles. (3) 2026 is incomplete — the file runs to 2026-09-09 and the count will rise through year-end, so every figure here is a year-to-date snapshot and dated as such. (4) The filter reads the manufacturer's language, so a genuine loss-of-power defect described only in other words is missed, and a campaign that mentions the phrase in passing is included. (5) A brand with more campaigns is not thereby less safe — recall counts also track fleet size, model breadth and how readily a manufacturer files. (6) Campaign-level only: nothing here describes an individual VIN.

Related checks

• How to check if a used Tesla has open recalls before buying
• Used Land Rover Defender problems to check before buying
• Used Jeep Grand Cherokee problems to check before buying

Sources

• NHTSA ODI Recalls flat file (FLAT_RCL_POST_2010.zip) — the dataset we computed over; retrieved 2026-09-12
• NHTSA Recalls flat-file data dictionary (RCL.txt) — field layout and date format
• Frozen filtered dataset (33 strict campaigns, CSV) — our published cut
• NHTSA recalls API — recallsByVehicle — per-campaign verification
• Ford Mustang Part 573 Safety Recall Report 26V547 (26C40) — ground-eyelet fracture / loss of drive power
• Electrek — Toyota recalls 8,521 electric C-HR SUVs — confirms 8,521 units and build dates
• Kelley Blue Book — Toyota recalls 2026 C-HR for potential loss of drive power
• Ford Authority — NHTSA defect petition DP26008 into ~500K 2015-2018 Ford Edge — context; investigation, not a recall
• NHTSA datasets & APIs landing page

Last verified: 2026-09-12