Why is the reported wait time at the border 15 minutes when you've been sitting in the line for an hour and a half? The answer is more interesting than "the government is bad at this" — it's about which line CBP is actually measuring, and the surprisingly old-school methods that produce most of the numbers. This guide walks through how US Customs and Border Protection (CBP) collects wait times, what each lane type means, why the numbers are sometimes way off, and how to use them anyway.
There are three ways CBP measures wait times — and most crossings use the worst one
Per CBP's own Wait Time Collection Methods document, estimated waits come from one of three techniques:
- Bluetooth sensing. Roadside readers detect the Bluetooth signals broadcast by phones, tablets, and laptops inside approaching vehicles. As each device is "seen" by readers further upstream and then again closer to the booth, CBP can compute how long the trip took. Reasonably accurate; deployed at most high-traffic crossings.
- RFID transponders (commercial vehicles only). Trucks in the FAST program carry transponders; readers along the commercial approach time-stamp them and compute travel time between checkpoints. Very accurate, but it only measures truck flow.
- Manual observation — line of sight and driver surveys. Officers eyeball the queue or ask drivers how long they waited. CBP calls this out explicitly: "All other crossings use a manual wait time collection method such as line of sight and/or driver surveys." Most smaller crossings rely entirely on this, and even sensor-equipped ports fall back to it when systems are down.
In the JSON feed we (Don't Wait) pull from CBP, the
automation field tells you which method is in use at each
crossing: BT for Bluetooth, RFID for trucks, or
blank/manual for the rest. Manual ports tend to round to 15- or 30-minute
buckets and update slowly — you'll see the same number sit there for hours
because no officer has filed an update.
What the lane types actually mean
Every CBP land port has multiple lane groups, each with its own reported wait. Picking the right one matters more than picking the right time of day:
Standard lanes (everyone)
The default — anyone with a passport, passport card, enhanced driver's license, or trusted-traveler card can use these. At busy crossings these are also the ones that back up first, so this is the wait time most travelers see and complain about.
Ready Lane (RFID-enabled documents)
A dedicated lane for travelers whose documents have an embedded RFID chip the officer's reader can pull biographical data from before you reach the booth — that's the time savings. Eligible documents include the U.S. Passport Card (not the regular passport book — the chip in those can't be read from far enough away), enhanced driver's licenses from Michigan, Minnesota, New York, Vermont, or Washington (or BC, Manitoba, Ontario, Quebec on the Canadian side), SENTRI cards, the new RFID Border Crossing Card, and the new Legal Permanent Resident "green card."
The key flexibility advantage: every adult (16+) needs some RFID-enabled document, but they don't all need to be the same kind. A driver with SENTRI and passengers with US Passport Cards can use the Ready Lane together. Kids under 16 can ride along on adult documents.
NEXUS (US ↔ Canada trusted-traveler)
Dedicated lanes only available to pre-approved NEXUS members at US/Canada land crossings. Every occupant of the vehicle must hold a valid NEXUS card — one non-member and the whole vehicle has to use the regular lane. The program is a joint US/Canada operation launched in 2002 after a pilot at Port Huron–Sarnia in 2000.
SENTRI (US ↔ Mexico trusted-traveler)
Dedicated lanes only available to SENTRI members at southern US land crossings. Two additional rules vs NEXUS: the vehicle itself must be separately registered with SENTRI (a one-time process), and every occupant of any age — including babies — needs a Trusted Traveler Program account. At San Ysidro the SENTRI vs. Standard gap can be two-plus hours.
FAST (commercial trusted-trader)
For pre-approved commercial trucks carrying pre-approved shipments. Both the driver and the carrier need to be enrolled. FAST lanes can move a shipment through commercial primary in minutes versus an hour-plus for unmodified trucks at peak.
Update cadence — slower than you think
CBP publishes wait times roughly hourly, not continuously. Even on Bluetooth-equipped crossings, the published number is an average of recent observations packaged for release on a cadence that's closer to every 60 minutes than every 60 seconds.
Practical implications:
- If a backup forms in the 20 minutes after the most recent update, the feed won't reflect it until the next push. That's how a "15 min" number can mean a 60-minute wait when you arrive.
- Conversely, when a backup clears (say, a lane just reopened), the feed can keep showing the old long wait for up to an hour.
- If a port is closed or unstaffed (overnight at smaller crossings), the feed shows "Lanes Closed" or "Update Pending" until someone resumes operations.
This is why our site re-polls every five minutes and also stores a rolling history — it catches changes within minutes of CBP publishing them, and the "best/worst times to cross" block on each crossing detail page surfaces the historical weekday-by-hour pattern that the current number can't tell you.
Five reasons the number is sometimes way off
1. The "wait" doesn't include the whole approach
CBP's clock starts at the point its sensors or observers can first detect you — typically a few hundred meters from primary inspection. Anything before that — toll plazas, surface streets backing up in Tijuana, traffic on I-5 itself — isn't counted. At San Ysidro the on-ramp backup alone can easily add 30+ minutes that the reported number ignores.
2. Manual ports round and lag
At smaller crossings, an officer is updating a number by hand. They round to coarse buckets (15, 30, 45 minutes) and don't always update when the line changes. A "30 min" reading from a small port should be read as "somewhere between 15 minutes and an hour, probably."
3. Bluetooth needs Bluetooth — and not everyone has it on
Bluetooth sensing depends on devices broadcasting discoverable signals as they pass readers. Older cars, phones with Bluetooth off, or trucks where the driver has paired their phone to the dash all show up to the sensor differently than a typical car. Sample bias can skew the measurement, especially during low-traffic periods.
4. Sudden events — enforcement, weather, lane closures
A vehicle pulled to secondary inspection, a sudden lane closure for staffing, a weather incident, or a high-profile enforcement event (drug/firearm seizure, etc.) can spike the actual wait in seconds. The feed will catch up at its next update cycle.
5. Holiday and event surges outrun the model
US Thanksgiving Wednesday, the Friday before a long weekend, the day after a major sporting event — these all produce wait patterns that diverge from the typical day. Bluetooth-based measurements eventually catch them, but the lag can be significant during the spike's first hour.
"Update Pending," "Lanes Closed," and other statuses
Each lane group in the CBP feed has an operational status field. The common values:
- Open — Normal operations, the lane is staffed and moving.
- Update Pending — The port hasn't reported a new number within CBP's expected window. Either the port is idle, the officer is busy with passengers, or a sensor is down. Not the same as "lanes closed."
- Lanes Closed — The lane group isn't operating right now (overnight, staffing, weather). At the lane-group level this is common; at the whole-port level it usually only happens at small crossings with hours.
- No Delay — Open and moving with effectively zero wait. You see this a lot at 3am on a Tuesday at outlying crossings.
We pass these statuses through to our card UI as-is — we don't try to "correct" or interpret them, on the principle that anything we made up would be less accurate than CBP's own report.
How to actually use wait time data
Given all of the above, here's how to get useful answers out of an imperfect dataset:
- Look at the historical pattern, not just the current number. Every crossing detail page on this site shows a "Best & worst times to cross" block computed from four weeks of snapshot history. That pattern is far more stable than the live number and tells you what to expect at the time you're planning to cross. Try Blaine Peace Arch or San Ysidro to see how it works.
- Compare lane-type waits side by side. The gap between Standard and NEXUS/SENTRI is bigger at peak times than off-peak. If you're a trusted-traveler member, the trusted-lane number is the one that matters — and it's often a fraction of the standard wait.
- Treat the current number as a floor, not a ceiling. "15 min reported" almost never means less than 15 minutes — but it can mean more, especially at manual ports or during the gap before the next update.
- Cross-reference with cameras when you can. Live cameras from WSDOT, Caltrans, and DriveBC show actual queues. If the reported number is 10 minutes but the camera shows traffic backed up a kilometer, trust your eyes. We embed available cameras on detail pages — see Sumas or BC Pacific Highway for examples.
A note on the Canadian side (CBSA)
The Canada Border Services Agency publishes its own wait times for northbound (US → Canada) traffic at the CBSA wait-times page. The structure is similar but the numbers come from a different system, so comparing "Canada → US" and "US → Canada" at the same physical bridge shows what's happening in each direction independently. CBSA publishes hourly during operating hours and shows "Not Applicable" for closed lane types. We mirror their values without modification.
Bottom line
The CBP wait-time feed is the best public dataset for this question, but it's a snapshot of CBP's own approximation — not a ground-truth timer. Use it for relative comparison (is Blaine quieter than Sumas right now? Is now a quieter hour than 30 minutes ago?), use the historical block on each crossing page for absolute planning, and use cameras when the decision actually matters. We make all three free.
See live waits and best-time charts for every US-Canada and US-Mexico land border crossing on the homepage, or browse our other guides for related background.