USA residential proxies: state, city and session guide

USA residential proxies route requests through home internet connections in the United States, so the target sees an address that a US consumer ISP such as AT&T, Comcast, Charter Spectrum or Verizon assigned to a household, on a network that IP databases classify as residential broadband. With state and city targeting, the exit can sit in the market whose prices, stock, search results or ads the job needs, rather than anywhere in the country.

A price check for a retailer returns a California total when the team needed Texas. In a local rank tracker, Chicago results pass for a national page. In both cases the exit was in the US, but the US is a stack of state and city markets, and a request has to land in the right one and then look like it belongs there.

What USA residential proxies are

Residential. A US residential proxy exits through a device on a home broadband line. The address belongs to an autonomous system (ASN) that a consumer ISP operates, so IP intelligence databases classify it as residential, not hosting. The residential proxy explainer covers how that classification works and how residential exits are sourced.

US. This address geolocates to the United States in the databases that sites use, and in most cases to a state and a city as well. That label is an estimate drawn from registry data, ISP feeds and measurement. MaxMind’s accuracy notes (checked September 2026) put it at 99.8% for country, about 80% for the US state and 66% for a US city within 50 km. Precision drops at each finer level.

Gateway. A customer never connects to the home device directly. Requests go to a provider gateway over HTTP or SOCKS5 with a username and password, and the gateway picks a US exit that matches the order’s location and rotation settings.

proxymint Lab #1, measured on 2026-09-28, sent its tests through a rotating residential package with a US exit and sticky rotation. Its exit sat on a home broadband line on AT&T’s network (AS7018) in Dallas, TX, and target hostnames resolved on the proxy side, through US public resolvers, when the client used HTTP CONNECT or socks5h. The Lab #1 results page has the method and the commands.

Why the state and city matter

A national exit is enough for a national page. Many US pages are not national.

Tax and checkout totals. Sales tax is set by states and localities. Five states, Alaska, Delaware, Montana, New Hampshire and Oregon, have no statewide sales tax, while California’s state rate alone is 7.25%, according to the Tax Foundation’s 2026 state and local sales tax table (2026). Collected from the wrong state, a cart total is a wrong data point even when the shelf price matches.

Stock and delivery. Retailers show availability, delivery windows and store pickup for the store or ZIP code they think the visitor is near, and the first guess comes from the IP address.

Local search. A query with local intent returns a map pack and organic results for the area around the searcher. From a New York exit, a tracker reading Houston rankings reads New York.

Ads. Campaigns are geotargeted by state, metro area and city, so verifying that a creative runs where it was bought means loading the page from inside that area.

US targeting levels explained

Every order for USA residential proxies picks a level. The finer the level, the smaller the set of online devices that can serve it, and the less certain the label the target reads.

On proxymint, rotating residential orders take country, region and city, and for the US the region is the state. Only places with devices online at the moment of ordering appear in the list, and the location can be changed later on a live package. For ZIP-level pages the reliable route is a city exit plus the site’s own store or ZIP picker, which is where the site keeps that setting anyway, and it holds even when an IP database labels the exit with a neighboring postcode.

How accurate US location labels are

State is often the best trade in the US. It carries most of what changes by market (tax, licensing, regional promotions), MaxMind places about 80% of US addresses in the right state, and a state draws on far more online devices than any one city in it.

City targeting earns its place when the page changes by metro area. Fewer households are online at any moment in a smaller city, on any residential network, so per-request rotation there repeats addresses sooner, and the request rate has to stay at what the target allows. If the city is not in the list, or the job needs more distinct addresses than one city returns, widen the order to the state, then to the country; the location can be changed on the live package.

Rotating or sticky US sessions

Rotation decides how long one US exit serves the job. proxymint’s residential orders offer three settings, chosen on the order and changeable later with the same logins: a new IP on every request, a timer from 5 to 60 minutes, or sticky for as long as the device stays online.

Every request. Fits stateless collection: product pages, search results, catalog checks across one state. Each request leaves from a different household, which spreads load across many addresses.

Timed, 5 to 60 minutes. Fits short flows that carry a cookie: a store selector, a cart, a few pages of a session. Pick a window longer than the flow.

Sticky. Fits a browser session that has to keep one address until it ends. Home devices go offline, so a sticky residential session lasts as long as that device stays online.

When one US identity has to hold for days or weeks, such as a seller dashboard or a partner’s firewall rule, a home connection is the wrong shape. That is the job of static ISP proxies: one US address assigned to the order for the whole term. The ISP vs rotating residential comparison walks through that split.

US use cases that fit

USA residential proxies serve jobs where the page differs for home users in a given market, or where a hosting address gets a different answer.

• Price and stock monitoring across states, including checkout totals with tax and local availability.
• Local search tracking by city, where the engine permits automated access.
• Ad verification on desktop and home networks, checking that a campaign runs in the states and metros it was bought for.
• Brand protection and marketplace monitoring, finding counterfeit or unauthorized listings as a US shopper sees them.
• Localization QA, confirming a US launch shows the right currency, shipping options and state-specific notices.

Legal and responsible use

Any automated access follows the same rules. Public pages only, the target’s terms first, and a block or a challenge means the job slows down or stops. A residential exit changes the network the request comes from; it does not change what the site allows.

Checking a US exit first

Before a job on USA residential proxies runs, confirm the exit is where the order says. Three requests settle it.

# 1. exit IP, network (ASN) and location as one database sees it
curl -s -x http://USERNAME:PASSWORD@HOST:PORT https://ipinfo.io/json

# 2. the same exit in a second database (put the IP from step 1 in place of EXIT_IP)
curl -s https://ipapi.co/EXIT_IP/json/

# 3. which resolver looks up names for this exit
curl -sL -x socks5h://USERNAME:PASSWORD@HOST:PORT https://edns.ip-api.com/json

Read the output this way:

• Country should be US in both lookups. Report a mismatch here; it is rare.
• Network should name a consumer ISP, not a hosting company or cloud provider.
• State should match the order in both databases. If the two disagree on city, the target may disagree too, and only the target’s own lookup decides.
• Resolver should be on the proxy side, not the client’s own ISP or office network.

Run this for a sample of exits, not one. Sampling a few dozen requests shows how many distinct addresses a state or city returns and how often the labels agree.

Keeping the session consistent with the exit

An exit IP is one location signal. Sites also read the browser’s time zone, language and DNS path, and a US exit with a Berlin clock is a mismatch any detection script can score. The guide to how websites detect proxies covers these checks in detail.

Time zone per state. The 50 states span six time zones, and Arizona (outside the Navajo Nation) and Hawaii do not observe daylight saving time, per the Bureau of Transportation Statistics (checked October 2026). A Phoenix exit should run with America/Phoenix, not America/Denver, which shifts by an hour every summer.

Locale. Set en-US so navigator.language and the Accept-Language header match a US visitor.

DNS. Use an http:// proxy URL or socks5h:// so names resolve on the proxy side. With plain socks5:// many clients resolve locally, and a CDN can answer for the client’s location instead of the exit’s.

In Playwright, the emulation options (checked October 2026) set locale and time zone per browser context, so each US market gets its own context:

const context = await browser.newContext({
  proxy: { server: 'http://HOST:PORT', username: 'USERNAME', password: 'PASSWORD' },
  locale: 'en-US',
  timezoneId: 'America/Phoenix',
});

Start each state or city in a fresh context, so cookies and a saved store from one market do not follow the job into the next.

Residential vs other US proxy types

Every proxy category has US jobs it fits. Pick by requirement.

Rotating residential proxies cover the broad consumer-origin work: many US addresses, chosen by state and city, rotated per request or held for a session. Mobile proxies fit apps and platforms that expect a carrier origin, with country, region and city targeting as well. Datacenter proxies handle volume on US targets that accept hosting networks, such as APIs and uptime checks.

Choosing USA residential proxies

Start from the page. If the page is the same across the country, a country-level US exit is enough and keeps the widest pool. If prices, tax, promotions or licensing change by state, target the state. Go down to city only when the page changes by metro area, such as local search results, store stock or a city ad buy, and set a ZIP inside the site when the page needs one.