Eurozone Inflation Rises to 3.8% as Energy Costs Climb

Eurozone inflation reached an estimated 3.8% in September 2026, up from 3.2% in August. Energy prices made the sharpest move among the main categories. For households, though, that shared headline is an average of bills that arrive separately. A rise in fuel or electricity costs can be felt now; further increases in the price of a meal or a service may take longer to appear, if they come at all.

Energy prices were 18.8% higher than a year earlier, compared with a 14.3% annual increase in August, Eurostat’s October 2 flash estimate shows. Annual services inflation rose from 3.0% to 3.2%, and food, alcohol and tobacco inflation rose from 1.1% to 1.4%. Inflation in non-energy industrial goods eased from 1.2% to 1.1%. Energy stands out in that list, but an annual rate cannot tell us on its own what happened to prices between August and September.

A Small Slice of the Basket, a Large Price Move

The figures come from the Harmonised Index of Consumer Prices, or HICP. Its annual rate compares September’s prices with those of September 2025. The monthly rate compares them with August 2026. Reading both prevents a year of price changes from being mistaken for a single month’s movement.

Energy prices rose an estimated 3.9% between August and September. Prices excluding energy rose 0.2% over the same month, while their annual inflation rate increased from 2.1% to 2.3%. The monthly energy move was much steeper than the change in everything else combined. Prices outside energy were rising too, though at a slower monthly pace when taken together.

Services prices fell an estimated 0.7% during September, although their annual inflation rate rose to 3.2%. The annual comparison depends partly on the price change a year earlier that drops out of the calculation. A customer pays September’s current price; the annual rate describes how it compares with last September’s. The monthly fall complicates any claim that September already shows an across-the-board acceleration. One monthly figure cannot rule out pressure within individual services or a later effect from energy costs.

Eurostat’s 2026 expenditure weights put services at about 46.8% of the euro-area index, against about 9.0% for energy. Energy’s smaller weight does not make an 18.8% annual price rise inconsequential: a large move in a smaller category can still pull up the headline, and the index weight is not any one household’s fuel or heating budget. Services occupy nearly half the measured basket. If businesses later pass higher energy costs into service prices, that pressure would extend across a much larger share of the index.

An energy inflation rate of 18.8% measures the change in energy prices from a year ago. It is not energy’s contribution, in percentage points, to either the 3.8% annual headline rate or its increase since August. The component’s weight and rate help explain why energy warrants attention, but those figures alone do not establish how much of the headline move it caused. That question also depends on the index calculation and changes elsewhere in the basket.

The rest of the basket resists a single label. Food, alcohol and tobacco account for about 18.9% of the index. Unprocessed food prices rose an estimated 1.3% from August, while processed food, alcohol and tobacco prices fell 0.2%. Non-energy industrial goods prices rose 2.1% during the month even as their annual inflation rate edged down. Someone buying fresh food or a manufactured good could therefore encounter a price movement unlike the services figure. These category readings measure price changes, not the share of each change caused by energy.

One Currency Area, Different Price Readings

September’s estimated annual inflation rate was 6.1% in Lithuania and 2.6% in Finland. The monthly figures produce another contrast: prices rose an estimated 1.8% in Greece and fell 0.4% in France. The euro-area headline describes prices across the currency union, but a national reading is more useful when the question is what happened to prices in a particular country.

Those comparisons still stop at the national average. Lithuania’s higher annual rate does not establish that its households faced a larger energy bill than Finland’s, just as France’s monthly decline does not establish that every French shopper paid less. A reader assessing exposure needs the local category changes and the mix of purchases that makes up their own budget. Neither a national headline nor the euro-area energy weight supplies that household-level picture.

Bulgaria joined the euro area in January 2026. Eurostat’s aggregate data through 2025 cover 20 countries, while 2026 data cover 21. Today’s currency area and the one measured during the inflation peak four years ago have different memberships. That is another reason to use the earlier episode to examine how an energy shock travels through prices, rather than treat its headline rate as a forecast.

What the Earlier Energy Shock Can Teach

Euro-area inflation reached 10.6% in October 2022, when energy directly contributed almost 4.5 percentage points, according to a SUERF analysis. Gas and electricity were the largest energy contributors during that episode, unlike earlier periods in which oil-linked fuels dominated. September 2026’s estimated 3.8% headline rate is far below the 2022 peak. The current flash figures give an energy inflation rate but no comparable fuel-by-fuel breakdown or direct contribution figure. They cannot establish that today’s shock has the same composition or will follow the same course.

The earlier episode also supplies a timing warning. In SUERF’s model-based analysis, the estimated contribution of energy-price shocks to inflation excluding energy reached its maximum in the first quarter of 2023. That is a modelled contribution across food, goods and services, rather than an observed services inflation rate. It illustrates a possible lag between an energy shock and price pressure elsewhere. It does not set a timetable for 2026.

The European Central Bank’s September staff projections anticipate a gradual rise in non-energy inflation until early 2027, while assessing indirect and further knock-on effects as contained in their baseline. Staff project average headline inflation of 3.0% in 2026 and 2.5% in 2027. Prepared before the October 2 flash estimate, that baseline is a forecast conditional on how energy prices and the wider economy develop, rather than a reading of September’s services bills.

The ECB also tests a severe scenario with a stronger, longer energy shock and a stronger response from wages and non-energy prices. It puts projected 2027 inflation at 5.4%, against 2.5% in the baseline, with weaker projected growth. The projections identify the Middle East conflict, the blockade of the Strait of Hormuz and volatile energy prices as sources of uncertainty. The gap between those paths turns on persistence and spillovers, questions a single flash estimate cannot settle.

On September 10, the ECB raised its three key interest rates by 25 basis points, taking the deposit rate to 2.50% from September 16. Its stated approach to further decisions is meeting by meeting, with attention to incoming data and underlying inflation. Higher borrowing costs can restrain spending and price pressure; they cannot produce more energy. The policy choice becomes harder if energy stays expensive long enough to feed into other prices while growth weakens. That is a risk described by the ECB’s scenarios, not an outcome demonstrated by September’s figures.

Eurostat plans to publish complete September figures on October 16. They may confirm or revise the estimated monthly fall in services and the much sharper rise in energy. After that, successive readings for services and prices excluding energy will help show whether the September energy jump remains concentrated or appears more widely in the bills households pay.