U.S. Retail Sales Fall 0.6% in July 2026, Largest Drop Since May 2025

U.S. retail and food services sales were $763.6 billion in July 2026, down 0.6 percent (±0.4 percent) from June and up 5.0 percent (±0.5 percent) from July 2025, the Census Bureau reported. It is the largest monthly move since May 2025.

The confidence interval is the part of that sentence to read carefully. A 0.6 percent decline with a ±0.4 percent interval is a decline the survey can actually distinguish from zero — which is not true of most monthly retail prints, and was not true of June's, which was unrevised at up 0.2 percent (±0.3 percent) and whose interval includes zero. July is a real move down, not sampling noise dressed up as one.

One caveat sits underneath every figure here: these estimates are adjusted for seasonal variation and for holiday and trading-day differences, but not for price changes. This is a nominal series. A nominal decline says nothing on its own about whether volumes or prices moved, and the release offers no price adjustment to separate them.

The Control-Group Read Is Softer Than the Headline

Stripping out the volatile components changes the size of the decline but not its direction:

• Total: down 0.6 percent
• Excluding motor vehicles and parts: down 0.3 percent, a change the survey flags as having a confidence interval that includes zero
• Excluding motor vehicles and parts and gasoline stations: down 0.2 percent, an interval that also includes zero

So the headline decline is concentrated in autos and gasoline, and the underlying consumer — the part of the series that feeds the consumption line of GDP most directly — was roughly flat rather than contracting. That is a materially less alarming reading than the headline, and it is the one to carry forward.

The three-month view is different again. Total sales for the May 2026 through July 2026 period were up 6.3 percent (±0.5 percent) from the same period a year ago. One soft month has not disturbed the annual trend, and the year-over-year figure at 5.0 percent remains firm.

For context on the excluding-autos measure, that series recorded its largest monthly move since January 2025 — the same story as the headline, one notch quieter.

Category Detail: Nonstore and Autos Carried the Decline

On a seasonally adjusted basis, the July move was concentrated in a few large categories:

• Nonstore retailers fell to $136.9 billion from $140.0 billion in June — the largest dollar decline of any category, and notable because this is the category that has carried retail growth all year, up 10.2 percent year-to-date
• Motor vehicle and parts dealers fell to $141.4 billion from $144.0 billion, and are up only 1.8 percent year-to-date, the weakest of the major categories
• Gasoline stations fell to $59.9 billion from $60.4 billion, though the category is up 14.9 percent year-to-date on price effects rather than volume
• General merchandise stores rose to $79.8 billion from $79.6 billion, and food services and drinking places rose to $103.6 billion from $103.0 billion

The pattern in that list is worth naming. Restaurants — the most discretionary line in the report — grew. General merchandise grew. The decline came from online and from autos, which are the two categories most sensitive to timing effects: promotional calendars in the case of nonstore, and inventory and incentive cycles in the case of vehicles.

Year-to-date, the category dispersion is wide. Miscellaneous store retailers are up 11.2 percent, nonstore retailers up 10.2 percent, and sporting goods, hobby, musical instrument, and book stores up 10.4 percent. At the other end, furniture and home furnishings stores are down 1.7 percent, food and beverage stores are up 1.0 percent, and motor vehicle and parts dealers are up 1.8 percent. Consumers are spending on goods bought online and on discretionary recreation while holding back on furnishing homes and buying cars — a distribution consistent with a housing market that is not turning over.

What a Single Advance Estimate Can and Cannot Tell You

These are advance estimates, drawn from a subsample of roughly 4,800 firms out of the more than three million retail and food service firms the full survey represents. They are computed with a link relative estimator rather than by imputing for nonrespondents, and they are benchmarked to the annual survey once it becomes available. The first print for any month is the noisiest one that month will receive.

Census has also flagged a broader restatement ahead: monthly retail sales estimates are expected to be revised for historical corrections and for the results of the 2023 and 2024 Annual Integrated Economic Survey, with revised estimates tentatively scheduled for September 28, 2026. That is a benchmark-style revision rather than a routine monthly one, and it can move the level and the growth path of this series together.

What to Watch

The August advance report publishes September 16, 2026, with the annual restatement following on September 28. Two things resolve July.

First, whether nonstore retailers rebound. A single month down after a year up 10.2 percent reads as a promotional-timing artifact; a second consecutive decline would mean the one reliable engine of retail growth this year has stalled, and no other category is large enough to replace it.

Second, whether the excluding-autos-and-gas measure stays flat or turns negative with an interval that excludes zero. July's version of that number is a statistical non-event — the survey cannot distinguish it from no change. If September's print shows the same measure down again and outside its interval, the softness has moved from the volatile categories into the core of consumer spending, and the 5.0 percent annual growth rate stops being the right frame.