FAA ATC delays are some of the most disruptive — and misunderstood — events in modern aviation. One minute you are pushing back from the gate on time. The next, the captain comes over the PA with that familiar phrase: "We've just been given a hold by ATC." Sometimes it's ten minutes. Sometimes it's three hours. And sometimes, as the entire country saw during the late-2025 government shutdown, the delays cascade across the whole national airspace system until thousands of flights are stuck on tarmacs from JFK to LAX.
So what are these delays, really? Who issues them, why, and how can you (as a pilot or a passenger) find out what's happening in real time? This guide walks through every type of FAA-issued delay, the systems behind them, and the language ATC uses when things get messy. If you've ever wondered why your flight is sitting at the gate while the weather looks perfect outside, you'll find the answer here.
The Federal Aviation Administration manages roughly 45,000 flights a day across U.S. airspace. That's a staggering number, and the system mostly works — until it doesn't. When demand exceeds capacity at any point in the network, ATC has tools to slow the system down. Those tools are what we collectively call "delays," and each one has a specific name and purpose. For more context on the agency that runs all of this, our FAA regulations page covers the full structure.
Imagine a freeway with a finite number of lanes. Now imagine that during a storm, two of those lanes close. Cars don't disappear — they just back up. The same thing happens in the sky. Runways, airspace sectors, and arrival corridors all have hard capacity limits. When weather, equipment failure, staffing shortages, or sheer volume push demand past those limits, the FAA slows the flow at the source rather than letting aircraft pile up in the air burning fuel.
That's the philosophy behind every delay program: hold the airplane somewhere safer and cheaper than the destination. Usually that means holding it on the ground at the departure airport. Sometimes it means rerouting it. Almost never does it mean canceling — cancellation decisions belong to the airline, not the FAA.
Not all delays are the same. Pilots see them on their flight plans as alphabet-soup codes; passengers usually just see "delayed" on the gate display. Here's what those codes actually mean, ordered roughly from least to most disruptive.
A GDP is the most common tool. It assigns each inbound flight a controlled departure time, sometimes called an EDCT (Expect Departure Clearance Time). The idea is simple: if the destination can only accept 50 arrivals per hour instead of the usual 80, the FAA spreads the inbound flights out so nobody has to hold in the air. You'll wait at the origin gate instead. Most GDPs last two to six hours. They're issued by the Air Traffic Control System Command Center in Warrenton, Virginia.
A ground stop is more severe. Instead of metering departures, ATC simply tells everyone bound for a specific airport: don't take off. It usually means something has gone very wrong — runway closure, severe thunderstorms over the field, a security incident, or a major equipment outage. Ground stops are typically short (under an hour) but can extend if conditions don't improve. Read our deeper guide on the FAA ground stop process for how they're declared and lifted.
Sometimes the problem isn't the airport — it's a chunk of airspace en route. Maybe a line of thunderstorms is sitting across the eastern seaboard, narrowing the usable corridors from six lanes to one. An AFP issues delays only to flights that would pass through the constrained area. If your route avoids that airspace, you go on time.
These are subtler. MIT tells controllers to space arrivals further apart than normal — say, 20 nautical miles instead of 10. The result is fewer flights per hour, which translates upstream into ground delays. Pilots rarely see MIT directly; it shows up as a tightened departure slot.
When convective weather builds in a known location (the New York metros are the classic example), the FAA pre-publishes a SWAP route. Aircraft fly the alternate route, which is usually longer. The extra distance burns more fuel and adds time, but it keeps everyone moving. SWAP isn't technically a delay program, but the rerouting feels like one if you're sitting in 24B.
Ground Delay Program — controlled departure times spread inbound traffic during reduced capacity. 2-6 hours typical, sometimes longer in major storms.
Ground Stop — all departures to a specific airport halted until further notice. Used for sudden severe events like thunderstorms or runway closures.
Airspace Flow Program — delays only flights routing through a constrained airspace block. Common during convective weather along the east coast.
Miles-in-Trail — increased spacing between arrivals reduces hourly throughput. Shows up upstream as tightened departure slots.
Severe Weather Avoidance Plan — pre-published reroutes around expected storm areas. Adds flight time but keeps traffic moving safely.
Expect Departure Clearance Time — the specific minute a flight is allowed to launch under a GDP. Windowed -5/+5 minutes around the assigned time.
Expect Departure Clearance Time. It's the slot the FAA gives your flight when a Ground Delay Program is active. Miss it by more than five minutes and you go to the back of the line. Hit it within the window (-5/+5 minutes) and you launch on schedule. EDCTs are issued by the FAA Command Center, propagated to airline dispatch systems, and shown in pilots' release packets. They are not negotiable in the short term, though Compression can pull them earlier when other flights cancel.
The FAA categorizes the root cause of every delay it issues. The breakdown is consistent year over year, though the proportions shift with the seasons.
Weather (70-75%). By far the biggest cause. Thunderstorms in summer, snow and ice in winter, low ceilings and fog in shoulder seasons. Weather doesn't have to be at your airport — a storm cell over Chicago can delay you in Boston because half the inbound traffic at BOS came through ORD airspace.
Volume (10-15%). Sometimes there's just too much demand. The morning push out of LGA, the evening bank at ATL — these are scheduled at or near the airport's hourly limit even on a perfect day. Add a 15-minute pause for anything and the whole bank slides.
Equipment (5-10%). Radar outages, radio failures, navigation aid problems. The 2023 NOTAM system crash that grounded every U.S. flight for hours falls in this bucket. So do the more routine "computer issues" that periodically pop up in announcements.
Staffing (5-10%). Air traffic controller shortages have been a chronic issue, especially since 2020. The FAA needs roughly 14,500 certified controllers but has consistently operated several thousand below target. The 2025 shutdown made this much worse — controllers worked unpaid, many called out sick, and the system buckled. Want to know more about how controllers work and earn? Our guide on the traffic controller career path has the numbers.
Runway construction or closures (under 5%). Self-explanatory but persistent.
FAA meteorologists predict reduced arrival capacity 2-6 hours ahead — typically due to weather or volume.
Command Center publishes the GDP with start time, end time, scope (which airports it covers), and target arrival rate.
Each inbound flight gets a controlled departure time. Airlines see it instantly via their dispatch systems.
Crews receive updated release times; passengers usually learn through gate announcements or airline apps.
If conditions improve faster than expected, the FAA cancels or shortens the program; flights get earlier slots.
You don't have to wait for the airline to tell you something is wrong. Several public-facing tools show exactly what the FAA is doing right now.
The single best source is fly.faa.gov, the Operational Information System. It lists active ground stops, GDPs, AFPs, and airport-specific delays with the official cause and expected duration. It's not pretty, but it's authoritative. Pilots use it daily.
For a more visual look, FlightAware and Flightradar24 both overlay weather, ATC programs, and live traffic. They're particularly useful when you're trying to figure out whether the storm cell in Pennsylvania is going to clip your route to Boston.
For pilots specifically, the FAA weather briefing tools and the official 1800wxbrief service include all current Traffic Management Initiatives. Adding a TFR check is also wise — temporary flight restrictions can cause delay-like routing surprises that aren't technically ATC delays but feel identical from the cockpit.
Airline apps are great for your specific flight but often slow to reflect FAA-wide programs. They'll show "delayed - air traffic" after the program is already half over. The FAA tools update within minutes of the program being issued.
The October-November 2025 government shutdown produced the worst ATC delay crisis in modern history. Controllers worked without pay for 35+ days. By week three, sick-call rates had tripled at major facilities. The FAA, lacking enough staff to safely run the system at normal capacity, imposed a 10% capacity reduction at the 40 busiest airports nationwide.
The result was predictable. Average daily delays exceeded 6,000 flights — roughly double the seasonal norm. Ground delay programs ran simultaneously at JFK, LGA, EWR, ATL, ORD, and DFW for days at a time. Some flights were delayed eight hours; thousands were canceled outright by airlines who couldn't reposition crews in time.
The lesson? ATC delays compound. A two-hour GDP at one hub doesn't just delay flights into that hub — it delays the next leg those aircraft were supposed to fly, and the next, and the next. By the end of a bad day, an airline can have 30% of its fleet out of position. Following the faa safety news during major events is genuinely useful because it gives you forward-looking signals airlines won't share.
Student and commercial pilots feel these delays differently. A GDP into a busy training airport can mean canceled lessons, lost solo time, and missed checkride windows. Worse, the practical test itself depends on ATC availability — if the local tower is short-staffed, your DPE may have to scrub. This is one of many reasons the FAA Airman Certification Standards include weather and ATC contingency questions in oral exams.
If you're working toward a certificate, plan a buffer into your timeline. Schedule your FAA web scheduler knowledge test for a mid-week morning when delays are statistically lowest. Get your FAA MedXPress medical done well before your checkride window opens. And keep checking the FAA flight delays dashboard the week of your test — if a major weather system is moving in, your DPE may want to flex the date.
Probably not soon. The FAA is hiring controllers as fast as the academy can certify them, but training takes 2-3 years per controller and attrition is high. Equipment modernization (the NextGen and now FAA Reauthorization 2024 programs) is rolling out, but its impact on delays is incremental. Meanwhile, passenger demand keeps growing.
The good news: technology is helping the demand side. Trajectory-Based Operations, Performance-Based Navigation, and time-based metering let the same airspace handle more flights more predictably. Data-Comm reduces voice congestion on busy frequencies. Surface management at airports (a.k.a. "Surface CDM") cuts taxi delays. So while bad days will continue to be bad, average days are slowly getting better.
If you fly often — for work, for training, or for fun — understanding what ATC is actually doing during a delay turns frustration into information. The system isn't picking on you. It's running the math, slowing the flow, and trying to land everyone safely. The next time you hear "we've got a hold from ATC," you'll know exactly what that means.
Here's something most travelers don't know: airlines build expected ATC delays into their schedules. A flight from LGA to ORD might be blocked at 2 hours 35 minutes on the timetable, even though the actual flying time is closer to 1 hour 50. That padding absorbs typical taxi waits, departure metering, and approach sequencing. It's why your aircraft sometimes "arrives early" — the schedule was conservative on purpose.
This padding helps on-time statistics but doesn't help travelers when a major program hits. Padded schedules assume average congestion. Once a GDP exceeds the average, the buffer evaporates and downstream legs run hot. That's why one bad afternoon at JFK can ripple into late-night cancellations at SFO. Crews run out of duty time. Aircraft are out of position. The bank collapses.
Sophisticated airlines (and increasingly the FAA itself) use Collaborative Decision Making — CDM — to share data on flight intentions in real time. When a carrier knows it will cancel a flight an hour before the slot, releasing that slot early lets another flight take it. The system gets a little more elastic. CDM is one of the quieter success stories in modern ATC, and it's been steadily reducing avoidable delays since the late 1990s.
From the cockpit, an ATC delay starts with a familiar message: "Expect departure delays of 90 minutes at this time. New EDCT is 1845 Zulu." The crew acknowledges, recalculates fuel, and may shut down one or both engines if the delay is long enough. Passenger comfort decisions get made in real time — sometimes the captain returns to the gate, sometimes they wait on the ramp with the APU running.
If the program is dynamic (capacity improving faster than expected), pilots watch for slot improvements. The FAA's Compression algorithm reshuffles slots when flights cancel or drop out, often moving everyone up by 10-30 minutes. If you've ever watched the departure board suddenly update to show your gate an hour earlier than promised, that's compression at work.
The flip side is slot extension. If conditions worsen, the FAA can push EDCTs further out. Pilots and dispatchers monitor advisories closely during this window. Knowing when to call for an updated release, when to swap aircraft, or when to recommend a cancellation is part of the operational art that doesn't show up in the airline's marketing.
FAA ATC delays aren't going away. But they aren't random, either. Every hold, every ground stop, every GDP has a reason — usually weather, sometimes volume, occasionally staffing or equipment. Knowing the codes, knowing the tools, and knowing where to look for real-time data turns you from a frustrated passenger or scrambling pilot into someone who actually understands what's happening in the world's busiest airspace system. Bookmark fly.faa.gov. Read the advisories. And next time the captain announces a hold, you'll already know whether it's a 30-minute hiccup or a four-hour problem.