COVID is growing in all 50 states, and the CDC said in June where it would start

The CDC's latest state-level estimate, built on data through Aug. 12, 2026, has COVID-19 infections growing or likely growing in all 50 states ^[1]. The useful part is not the number 50 but the fact that the agency called the shape of this in June, roughly two months earlier ^[6][30], which moves the summer wave out of the surprise category and into the forecastable one.

The weekly summary highlights increases across the South and West, while the 50-state estimate shows infections also rising in the Northeast and Midwest ^[3][4]. Activity remains low in many regions; the direction is what has changed ^[5].

That regional split is what the June 2026 outlook anticipated: more summer transmission in Southern and Western areas that saw comparatively little COVID-19 activity during the 2025-2026 winter, on the reasoning that those regions now have less recent immunity from infection ^[6][7]. The Northeast and Midwest, which had more winter activity, were expected to see smaller summer increases ^[8]. The agency treats the explanation as tentative ^[9]. COVID-19 can peak in summer or in winter, with timing varying by region, and researchers are still working out how waning immunity, new variants, behavior, and environmental conditions combine in any given wave ^[9][10].

The measurement stack matters here. State estimates come from emergency department data, with week-to-week changes in transmission modeled to produce R(t), an estimate of how many additional infections arise on average from each infected person ^[11][12]. Wastewater is a separate and earlier signal: labs detect viral material shed by people with and without symptoms, sometimes before rising infections appear in clinics and hospitals, though it can only show whether the virus is becoming more common, not how many people are infected ^[13][14]. Traditional case counts have degraded as home testing has grown and results go unreported, so the agency triangulates emergency department visits, lab results, hospital data, and wastewater ^[15][16].

Transmission and severe outcomes are not moving together. National hospitalization peaks have declined since the 2023-2024 winter, which the CDC attributes to some combination of greater population immunity, the absence of a dominant variant with pronounced immune escape, and possibly lower severity ^[17][18]. That is a statement about peaks, not about individuals: the agency notes growing transmission still carries serious risk for older adults and others more likely to become severely ill ^[19][20].

Variant data lags the transmission picture. The most recent posted estimate has XFG.1.1 at 29.7 percent of circulating variants for the two weeks ending July 4, 2026 ^[21], about 39 days behind the Aug. 12 transmission cutoff ^[31], with the remaining roughly 70 percent spread across other lineages ^[32]. The FDA has selected an XFG strain for the 2026-2027 vaccine, intended for use beginning this fall ^[22].

Two operational notes for anyone running a workplace policy. A single negative home antigen test can miss an early infection; the FDA recommends symptomatic people retest after 48 hours and asymptomatic people take three tests 48 hours apart, a four-day protocol ^[23][24][33]. Treatment windows are short: Paxlovid must start within five days of symptoms, remdesivir within seven, and the FDA-approved post-exposure option Xocova (ensitrelvir), for ages 12 and up, within 72 hours ^[25][26]. CDC guidance is to stay home until symptoms have improved for 24 hours with fever gone without medication, then consider added precautions for five more days ^[27].