Weekly estimates of retail gasoline and diesel prices, observed on Mondays, by state and region. Use the fuel selector to switch between them, and the price selector to switch between inflation-adjusted prices, in 2024 dollars, and nominal prices — the dollars of the day. Prices reflect the cost of crude oil, refining, distribution, and local taxes, and vary considerably across states. To see the two fuels side by side, use the gasoline vs. diesel chart.
Each weekly figure is a snapshot, not an average over the week. EIA surveys retail outlets for the cash pump price — self-serve, including taxes — as of 8:00 a.m. local time on Monday, and publishes the estimate the following morning. The figure is a volume-weighted average across outlets at that moment, weighted by their annual sales volumes: about 1,000 outlets for gasoline (Form EIA-878) and 590 for diesel (Form EIA-888). So a price shown for a given week is that Monday’s price, and moves within the week do not appear. Gasoline is regular grade, all formulations; diesel is No 2 diesel, all types. State-level data are available for states with sufficient retail coverage; other states are grouped into regional aggregates. EIA publishes diesel for fewer areas than gasoline — the US and the PADD regions, but among individual states California alone, and no cities — so selecting diesel narrows the list, and a state selection that has no diesel series is dropped.
Coverage starts at different dates by area. Gasoline: the US from August 1990, the five PADD regions from May 1992, California from May 2000, and the remaining states between 2000 and 2003. Diesel: the US and the PADD regions from March 1994, California from July 1995.
There are two methodology breaks in the series. EIA rebuilt its sample and estimation method for gasoline on 14 May 2018 and for diesel on 13 June 2022, and says estimates either side of those dates are not directly comparable. In practice the discontinuity is small: prices moved 2.8 cents across the gasoline break and 1.5 cents across the diesel one, which is around the median weekly move for each fuel, so it is not distinguishable from ordinary week-to-week variation. EIA did not publish both old and new estimates for the same week, so the size of the methodology effect itself cannot be separated from genuine price movement. It is worth knowing about for comparisons that reach across those dates — such as anything measured against the 2008 peak — but it is far too small to affect them.
The diesel product itself changed over the period. The “all types” series carried low-sulfur diesel through the switch to ultra-low-sulfur diesel; from mid-2010 it is fully represented by ULSD, and the two series are identical thereafter. This does not invalidate the long run, but prices before and after the switch are not quite like for like.
The bar and table views show an annual figure: the unweighted mean of that year’s weekly Monday readings. It is not EIA’s own annual average, which weights by sales volume through the year.
To adjust for inflation, nominal prices are deflated using the monthly seasonally adjusted CPI-U (all items, all urban consumers — FRED series CPIAUCSL), rebased to 2024 dollars. BLS publishes a month's index about two weeks after the month ends, so the most recent weeks are deflated with the latest published month and revised once the index catches up; the nominal view is unaffected, being EIA's published figures as they stand.