The news: Amazon reported second-quarter results showing $27.9 billion of shipping costs, up 19% from a year earlier, according to its quarterly filing. The number of items customers bought grew 17%.
Shipping costs outgrowing units means the average cost of moving one item went up. That had not happened at Amazon since the second quarter of 2022.
CFO Brian Olsavsky did not blame the delivery network. The causes, he said on the earnings call, were “fuel inflation from the conflict in the Middle East and higher line haul rates from driver capacity limitations.”
By the numbers: Amazon publishes both growth rates every quarter in its earnings release. The cost-per-item figure here comes from comparing them, so it is derived rather than reported. It is also rough, because paid units is a worldwide count that includes seller-fulfilled and digital items Amazon does not ship.
With that caveat, the two series lined up back to 2022 show a 15-quarter run in which units grew at least as fast as shipping costs. One of those quarters, the third of 2023, was a tie rather than a saving.
The gap had been narrowing for a year before it closed. It ran about 5.4 points in Q2 2025, then 2.7, then 1.8, then 0.9, and now sits roughly 2 points the other way.
Because the crossing had been coming for a year, Olsavsky could argue the network itself is still getting more efficient underneath it. “Excluding the impact of higher fuel and line haul rates, shipping costs grew more slowly than worldwide unit growth at a pace that is relatively consistent with last quarter,” he said.
Sellers also picked up part of the bill. A 3.5% fuel and logistics surcharge on Fulfillment by Amazon fees, about 17 cents on an average item, took effect in April. That surcharge partially offsets the higher transportation costs, Olsavsky said.
The pattern: Amazon was late to this. Five other large freight buyers and carriers reported the same fuel shock before Amazon did, and they sit on opposite sides of the invoice.
The carriers collected the cost and passed it on:
UPS fuel expense rose 60.4% in the June quarter, while surcharges brought in roughly the same amount. “So it really had a minimal impact on profit,” CFO Brian Dykes said.
FedEx fuel costs rose 70% in its Federal Express segment in the fiscal quarter that ended in May, and surcharges supplied 5 points of a 14% revenue gain, per Supply Chain Dive reporting.
FedEx Freight, now a standalone company, grew revenue per shipment 11.5% that same quarter on base rates its CFO called substantially flat, so most of that gain was fuel.
The shippers paid it:
Chewy booked a low-single-digit-million hit from carrier fuel surcharges in its first quarter and guided to a mid-single-digit-million headwind in the second.
Walmart absorbed “approximately $175 million or about 250 basis points of operating income growth from higher-than-planned fuel costs,” CFO John Rainey said.
Net fuel surcharges on ground parcels ran 40% above last year in the second quarter, per Supply Chain Dive, so anyone shipping parcels paid a version of this.
Amazon belongs on both lists, because it buys transportation like a shipper and sells it like a carrier. This quarter it did both: it carried the fuel cost in its own network and billed sellers a surcharge to recover part of it.
On-highway diesel, the input under every one of those numbers, averaged $5.35 a gallon over the second quarter, against $3.56 in the same months of 2025, a 50% increase. Truckload capacity tightened over the same period, amid federal enforcement of English-proficiency rules for drivers, non-domiciled CDL restrictions and a wave of carrier exits. That is why Olsavsky named driver capacity as his second cause.
Most of that fuel move came early. Diesel rose $1.17 a gallon in the first half of March, Amazon announced its surcharge in the first days of April, and the surcharge took effect on April 17. Walmart and Chewy have fee levers of their own, but neither moved one that fast at that scale.
The counter: The streak may not have ended at all, because the number behind it is soft. A two-point move in a figure built from two rounded, differently constructed disclosures is small enough that product mix alone could account for it.
Nothing in the filing separates “the cost of shipping an item rose” from “the share of counted units Amazon actually ships fell.”
Prime Day is the other live explanation, and it is at least as plausible as diesel. The event ran June 23 to 26 this year, inside the second quarter, after running July 8 to 11 last year.
Concentrating that much demand into a few days raises cost per item through surge labor, split shipments and less consolidated line haul, without anything changing in the underlying operation.
The pass-through cuts the wrong way for Amazon too. The 3.5% surcharge moves part of the cost onto third-party sellers, who put it into prices, but cost per item still rose after that. So the surcharge makes the efficiency story weaker rather than stronger.
And the strongest evidence that the network is fine comes from Amazon itself. The ex-fuel comparison is the company’s own, and the cost-to-serve figure is company-defined and unaudited. The fuel adjustment also excludes line haul rates, which depend partly on how much line haul Amazon chooses to run.
What’s next: Amazon has run this play before. It added a 5% fuel and inflation surcharge in 2022, then folded it into standard fulfillment rates the following year and removed the separate line.
Whether this one hardens the same way is the question for the third-quarter report in late October. That print also drops Prime Day out of the year-over-year comparison, so if cost per item is still rising without it, fuel is not the whole explanation.
