Decision sight distance is the length of roadway a driver must be able to see ahead to detect an unexpected condition, understand it, choose a response, and complete that response safely. AASHTO’s Green Book defines five avoidance maneuvers, labeled A through E, with required visibility ranging from 220 feet at 30 mph on a simple rural road up to 1,445 feet at 70 mph in a complex urban environment. The calculation is a design speed multiplied by a pre-maneuver time (and, for stopping cases, a braking term), giving distances well in excess of ordinary stopping sight distance because they credit the driver with the seconds actually spent scanning, filtering, and deciding before any input to the brake or steering wheel.
Why It’s Longer Than Stopping Sight Distance
Stopping sight distance assumes a clean scenario: a driver sees an object in the lane and brakes. The standard uses a 2.5-second perception-reaction time, which works when the hazard is obvious and the only sensible response is to stop. Decision sight distance exists for every situation that doesn’t fit that model. A driver approaching an interchange who has to work out which lane feeds the correct ramp, or who meets a lane drop with little warning, needs more than 2.5 seconds to think.
The pre-maneuver times built into decision sight distance run from 3.0 seconds for the simplest case up to 14.5 seconds for the most demanding one. At 60 mph, stopping sight distance on a level road is roughly 570 feet. Decision sight distance at the same speed ranges from 610 to 1,280 feet depending on the maneuver, so the hardest urban scenarios need more than twice the visibility of a basic stop.
The Five AASHTO Avoidance Maneuvers
The Green Book organizes decision sight distance around five maneuver types reflecting increasing complexity.
- Maneuver A, stop on a rural road. The driver detects a hazard and brakes to a stop where traffic density and visual clutter are low. Pre-maneuver time is 3.0 seconds.
- Maneuver B, stop on an urban road. The same braking response, but in an environment with far more competing information from signs, signals, adjacent traffic, and commercial development. Pre-maneuver time jumps to 9.1 seconds because drivers take longer to isolate the actual hazard from the visual noise.
- Maneuver C, speed, path, or direction change on a rural road. Instead of stopping, the driver must steer around the hazard or change lanes on a lower-traffic road. Total time ranges from 10.2 to 11.2 seconds depending on speed.
- Maneuver D, speed, path, or direction change on a suburban road. The same evasive action in a suburban corridor with more intersections, driveways, and traffic. Total time ranges from 12.1 to 12.9 seconds.
- Maneuver E, speed, path, or direction change on an urban road. The most demanding category. The driver must change speed or direction in a dense urban environment with heavy signage, adjacent buildings, and competing traffic movements. Total time ranges from 14.0 to 14.5 seconds.
A and B involve stopping; C, D, and E involve steering, lane-changing, or speed adjustment. The split matters because the two groups use different formulas.
How Decision Sight Distance Is Calculated
AASHTO uses two formulas depending on whether the driver stops or steers.
Maneuvers A and B (Stopping)
For maneuvers where the driver brakes to a halt, the formula adds a pre-maneuver travel distance to a braking distance:
DSD = 1.47Vt + 1.075(V² / a)
V is the design speed in miles per hour, t is the pre-maneuver time in seconds (3.0 for Maneuver A, 9.1 for Maneuver B), and a is the assumed deceleration rate in feet per second squared. The factor 1.47 converts mph to feet per second. The braking component works the same as in stopping sight distance; the pre-maneuver time is what makes the distance grow.
Maneuvers C, D, and E (Path or Speed Change)
When the driver steers rather than brakes, the formula drops the braking term and uses only a time-based distance:
DSD = 1.47Vt
Here, t is the combined pre-maneuver and maneuver time. The maneuver component itself runs between 3.5 and 4.5 seconds and actually decreases as speed increases, because at higher speeds drivers tend to make smoother, faster lane changes rather than sharp corrections. Total time still grows with complexity from C through E, which keeps the urban values the highest.
Values by Design Speed
The following table gives the decision sight distance values in feet from the Green Book (Table 3-3). These are the numbers designers work from when laying out a segment where decision sight distance applies.
| Design Speed (mph) | A (ft) | B (ft) | C (ft) | D (ft) | E (ft) |
|---|---|---|---|---|---|
| 30 | 220 | 490 | 450 | 535 | 620 |
| 40 | 330 | 690 | 600 | 715 | 825 |
| 50 | 465 | 910 | 750 | 890 | 1,030 |
| 55 | 535 | 1,030 | 865 | 980 | 1,135 |
| 60 | 610 | 1,150 | 990 | 1,125 | 1,280 |
| 65 | 695 | 1,275 | 1,050 | 1,220 | 1,365 |
| 70 | 780 | 1,410 | 1,105 | 1,275 | 1,445 |
A few patterns are worth noting. Maneuver A values are the lowest at every speed because a rural stop carries the least cognitive load. Maneuver B values are surprisingly high and often exceed C and D, because even though the driver is just stopping, the urban environment demands enough extra processing time (9.1 seconds) to balloon the distance. At 70 mph, a Maneuver B scenario calls for nearly a quarter-mile of unobstructed visibility.
Eye and Object Height Assumptions
Every sight distance calculation in the Green Book assumes a driver eye height of 3.5 feet and a target object height of 2.0 feet, representing the tail lights of a vehicle ahead. These are 5th-percentile values, so 95 percent of vehicles on the road have taller driver positions and higher-mounted tail lights than the design assumes.1Federal Highway Administration. Speed Concepts: Informational Guide – Chapter 4: Engineering and Technical Concepts Designing to the shortest plausible driver and the lowest plausible tail light keeps the geometry workable for nearly everyone on the road.
These heights matter most on vertical curves, where the crest of a hill can block the line of sight between a low-seated driver and a low-mounted object. If the crest is too sharp, the available sight distance drops below the required value and the curve must be flattened or lengthened.
Where It Applies
Decision sight distance is not required everywhere. Ordinary road segments use stopping sight distance as the baseline. AASHTO describes decision sight distance as desirable at locations where drivers are likely to make errors in information processing, decision-making, or vehicle control, and the Green Book calls out three categories of problem spots.
- Interchanges and intersections with unusual geometry. Exit ramp gore areas, left-side exits, and any location where the required maneuver isn’t what drivers expect. Most drivers instinctively look right for exits, so spotting a left-side exit takes significantly more processing time.
- Cross-section changes. Toll plazas, lane drops, and transitions from highway to local road. A four-lane highway that suddenly becomes two lanes forces a merge decision under time pressure, and drivers who don’t see it coming make abrupt lane changes.
- High visual noise areas. Stretches where roadway elements, other vehicles, traffic signals, and advertising signs all compete for attention. Commercial corridors near freeway interchanges are the worst offenders, because the driver has to find a specific sign or lane assignment while filtering out dozens of irrelevant inputs.
Designers are also instructed to avoid placing intersections within a “reduced decision zone,” a stretch where stopping sight distance exists but decision sight distance does not. If an intersection falls within such a zone, the preferred fix is to relocate the intersection or adjust the road’s vertical profile to restore full decision sight distance.
When Full Distance Can’t Be Provided
Terrain, existing structures, and right-of-way constraints sometimes make the table values impossible to achieve. AASHTO does not treat this as an automatic failure. The Green Book directs designers to compensate with advance warning through traffic control devices: overhead lane-assignment signs placed well upstream, advance guide signs for exits, or flashing beacons alerting drivers to an upcoming decision point. The goal is to shift the detection and decision phases earlier in the approach so the remaining geometric distance is enough for the maneuver itself.
That flexibility makes decision sight distance function more as a design target than an absolute minimum. A road can be safe without meeting the full value if the information gap is addressed another way. When a project has the room and budget to provide the full distance, though, there is no good reason not to. Advance signs can be obscured, ignored, or vandalized; geometric sight distance is always there.
Grade, Friction, and Obstructions
Road grade affects how quickly a vehicle can decelerate, which affects the braking portion of the calculation. On a downgrade, gravity works against braking and lengthens stopping distance; on an upgrade, gravity shortens it. The Green Book notes that sight distance available on downgrades tends to be naturally longer than on upgrades, which roughly offsets the braking penalty without an explicit correction. For that reason, many designers don’t adjust decision sight distance for grade at all. The main exception is one-way roadways or divided highways with independent profiles, where the automatic offset doesn’t apply and grade corrections may be warranted.1Federal Highway Administration. Speed Concepts: Informational Guide – Chapter 4: Engineering and Technical Concepts
Pavement friction matters for the braking component of Maneuvers A and B. Wet or icy surfaces reduce the effective deceleration rate, which extends the braking distance. Designers working in regions with frequent ice or heavy rain sometimes use a lower deceleration assumption, though AASHTO’s published values assume a standard wet-pavement friction coefficient.
Sight line obstructions are separate from the calculation. Vegetation, retaining walls, bridge piers, and large signs can block the view ahead, effectively reducing available sight distance regardless of what the geometry would otherwise provide. A segment that calculates to 1,280 feet of decision sight distance delivers less than that in practice if a stand of trees sits inside the line of sight.