In the United States, we switch to daylight saving time on the second Sunday in March, at 2:00 a.m. local standard time, when clocks spring forward one hour. In 2025 that date is March 9; in 2026 it is March 8. The switch back to standard time happens on the first Sunday in November, at 2:00 a.m. daylight time, when clocks fall back one hour.
The mechanism: why 2:00 a.m. and why Sunday
The clock change is not triggered by sunrise, temperature, or any natural event. It is a legal schedule written into federal statute and executed by state and local governments. The precise trigger is the calendar rule in the applicable law: the second Sunday of March and the first Sunday of November for most of the country. The 2:00 a.m. anchor is a practical choice, not a scientific one.
Why 2:00 a.m.? It is the hour with the lowest traffic volume and the fewest commercial transactions on a weekend night. A one-hour jump at that hour minimizes disruption to transportation schedules, payroll systems, and broadcast programming. The instant itself is defined as 2:00 a.m. local standard time on the spring date; the clock skips directly from 1:59:59 to 3:00:00. In fall, 1:59:59 daylight time becomes 1:00:00 standard time, so the 1 a.m. hour repeats.
Sunday was chosen because it precedes the Monday workweek, giving institutions one day to reconcile clocks, schedules, and software before normal business resumes. The result is a coordinated, nationwide shift that most people experience as a lost or gained hour of sleep rather than a logistical event.
Quantified context: the hour, the dates, the spread
The shift moves one hour of clock time, but the practical effect is a redistribution of daylight across the waking day. In spring, sunrise and sunset both appear one hour later on the clock. In fall, both appear one hour earlier.
The U.S. schedule currently produces roughly 238 days of daylight time and 127 days of standard time per year, because the March-to-November window is longer than the November-to-March window. That asymmetry is deliberate: Congress extended daylight time in 2007, moving the spring start from the first Sunday in April to the second Sunday in March and the fall end from the last Sunday in October to the first Sunday in November. The change added about four weeks of daylight time per year.
The date of each switch moves every year because it is tied to a weekday, not a fixed calendar date. The spring switch falls between March 8 and March 14. The fall switch falls between November 1 and November 7. That is why the answer to the question changes annually, and why calendars, phones, and scheduling software must be updated rather than assumed.
Not every state observes the same rule. Arizona and Hawaii do not observe daylight time. So do Puerto Rico, Guam, American Samoa, and the U.S. Virgin Islands. Their clocks stay on standard time year-round, which means the local time difference between them and daylight-time states shifts by one hour twice a year.
Timeline: a typical switch weekend
| Time (local) | Spring forward | Fall back |
|---|---|---|
| Saturday evening | Clocks still on standard time | Clocks still on daylight time |
| 1:59 a.m. Sunday | Standard time, final minute | Daylight time, final minute |
| 2:00 a.m. Sunday | Clocks jump to 3:00 a.m. | Clocks drop to 1:00 a.m. |
| 1:00–1:59 a.m. | Does not exist that night | Occurs twice |
| Sunday morning | Sunrise appears one hour later | Sunrise appears one hour earlier |
| Monday | First full workday on new time | First full workday on new time |
| Following week | Body clock adjusts gradually | Body clock adjusts gradually |
The spring gap is the reason some digital systems log errors: a timestamp of 2:30 a.m. on that Sunday is not a valid local time, and naive software can misorder events. The fall overlap is the mirror problem: 1:30 a.m. occurs twice, so logs need a time-zone offset to disambiguate.
Edge cases that break the simple rule
The federal rule is a default, not a universal law. Several categories of exception matter.
States and territories that opt out entirely: Arizona (except the Navajo Nation), Hawaii, Puerto Rico, Guam, American Samoa, and the U.S. Virgin Islands stay on standard time year-round. States that have considered permanent daylight time: several have passed legislation contingent on neighboring states or federal approval, so their current status can differ from their announced intentions. Border and travel effects: crossing into a non-observing state or territory changes the local hour offset, which can surprise travelers and logistics software. Devices and scheduled systems: phones usually update automatically, but medical devices, industrial controllers, and older scheduling systems may not. Workplace schedules: employers, not the federal government, decide how shifts and pay periods align with the change.
A counterintuitive detail: the spring switch is not the day most people feel worst. Sleep researchers generally find the adjustment is gradual, and the effects accumulate over the following days rather than peaking on Sunday itself. The fall switch, which adds an hour, is often assumed to be harmless, but it also disrupts sleep timing and can produce its own adjustment period. The direction of the change does not map neatly onto how people feel.
Misconceptions worth correcting
The most common misconception is that daylight time exists to help farmers. It does not. Agricultural schedules have historically been shaped by sunrise and season, not clock time, and farm groups have often opposed the change. The modern rationale is a mix of energy policy, commerce, and convenience, and the energy savings are modest and debated.
A second misconception is that the switch is universal in the United States. It is not, as the exceptions above show. A third is that the dates are fixed. They are not; they are calendar rules that move each year.
A fourth, subtler misconception concerns the clock itself. The switch does not change the length of the day or the amount of sunlight. It changes the label on the hour. The sun rises and sets at the same physical times; only the numbers on the clock move. That distinction matters for anyone reasoning about sunrise, sunset, and outdoor light.
Practical actions and thresholds
There is no official threshold that says when to prepare, but a few concrete steps reduce friction.
Set a reminder for the Saturday before each switch. The spring date falls between March 8 and March 14; the fall date between November 1 and November 7. Adjust bedtime by 15–20 minutes on the two nights before the spring switch. This spreads the one-hour shift across a few days rather than forcing it in one night. Check non-automatic devices: wall clocks, car clocks, medical devices, thermostats, and industrial controllers. Confirm scheduling software handles the invalid 2:00–2:59 a.m. spring interval and the repeated 1:00–1:59 a.m. fall interval. For travel or cross-state work, verify whether the destination observes daylight time before assuming the hour offset.
Limitations and what the schedule does not promise
The schedule answers the when, not the whether. Federal law sets the default dates, but Congress can change them, and states can opt out. The current rules reflect the 2007 extension, and proposals to make daylight time permanent or to abolish the switch have circulated for years without a stable nationwide change. Any answer about future dates should be checked against the current statute.
The scientific picture is also limited. Research on the health and safety effects of clock changes is mixed, and the size of any effect depends on latitude, occupation, and individual sleep patterns. For most people the change is an inconvenience rather than a hazard, but the evidence does not support strong universal claims in either direction.
The practical rule is simple: the switch happens on the second Sunday in March and the first Sunday in November, at 2:00 a.m. local time, for most of the United States. Everything else, from state exceptions to software behavior, is a detail that the calendar rule does not capture.