Skip to main content
Category: Laws
Type: Psychomotor / human–computer interaction model
Origin: Paul M. Fitts, Journal of Experimental Psychology (1954)
Also known as: Fitts’ law; pointing time model
Quick Answer — Fitts’s Law predicts that the time to point to a target rises with the distance to the target and falls as the width of the target grows—roughly as a logarithm of their ratio. Paul Fitts published the model in 1954; HCI later used it to compare devices and place controls. Design for frequent actions means larger, nearer targets—or infinite edges that remove overshoot cost.

​
What is Fitts’s Law?

Fitts’s Law is a predictive model of aimed movement: the average time to move rapidly to a target depends on the ratio of distance to the target to target width along the movement axis.
The farther and smaller the target, the longer the point—logarithmically, not as a flat “twice as far, twice as long.”
Think of parking a car in a garage: a wide bay close by is easy; a narrow bay far away forces a long approach and a careful final crawl. Fitts quantified that speed–accuracy tradeoff for tapping, transferring, and, later, screen pointing. In the common form, movement time is MT = a + b · ID, where the index of difficulty ID grows with log₂(2D/W) in Fitts’s original setup (or log₂(D/W + 1) in the Shannon form popular in HCI). Constants a and b depend on the limb, device, and conditions. The law sits beside Hick’s Law (choice among alternatives) and compresses ratios in a spirit related to Weber–Fechner scaling—here for motor “information,” not sensation intensity.

​
Fitts’s Law in 3 Depths

  • Beginner: Big, nearby buttons feel “easy to hit”; tiny, distant ones invite misses and slowdowns.
  • Practitioner: For each primary action, shrink D (place near the prior focus) and grow W (size and hit area); put critical mouse targets on screen edges or corners when the pointer cannot overshoot.
  • Advanced: Treat ID as a budget for motor information under a speed–accuracy tradeoff; device throughput and edge geometry change a and b, and touch edges do not behave like mouse walls.

​
Origin

Paul Morris Fitts (1912–1965) published “The information capacity of the human motor system in controlling the amplitude of movement” in June 1954 in the Journal of Experimental Psychology (Vol. 47, No. 6, pp. 381–391). Drawing on information theory (Shannon-era ideas of channel capacity), he framed aimed movement as transmitting bits: distance like signal amplitude, tolerance (width) like allowable noise. Classic tasks included reciprocal tapping between plates, disc transfer, and pin transfer. He defined an index of difficulty and an index of performance (bits per second)—today often called throughput. HCI adopted the model for virtual pointing. I. Scott MacKenzie popularized the Shannon formulation ID = log₂(D/W + 1) in the early 1990s and, with William Buxton, extended analysis to two-dimensional targets (CHI ’92). ISO 9241 later referenced Fitts-style measures for input-device evaluation. Practical UX writing (for example Nielsen Norman Group summaries) translates the math into placement and sizing rules for menus, buttons, and edges.

​
Key Points

Fitts’s Law is a motor budget, not a license to make every control huge. Use it when pointing cost is part of the task.
1

Distance and width jointly set difficulty

Difficulty tracks the ratio D/W, not either variable alone. Halving distance or doubling width both ease the task; tiny far targets compound both penalties. A 44-pixel icon across the screen can feel harder than a larger control beside the cursor.
2

Logarithmic growth, two movement phases

Time does not scale linearly with distance: an early ballistic phase covers ground; a slower corrective phase lands on the target. Small W mainly inflates that final “parking” cost—why cramped icons feel exhausting even when they are not far.
3

Edges and corners as infinite targets (pointer UIs)

On a mouse-driven display, the cursor stops at the screen edge. Targets flush with an edge act as infinitely deep along that axis: users can fling the pointer without overshooting. Corners join two edges—“magic corners”—which is why system menus and Start-style controls often live there.
4

Measure device and task, not slogans

Fitts parameters and throughput let labs compare mice, trackpads, and keys under controlled D–W sets. Pair results with Hick’s Law when option count, not only pointing, drives latency—and with error rates, not movement time alone.

​
Applications

Use Fitts wherever a finger, cursor, or hand must acquire a control under time pressure.

Desktop and OS chrome

Place frequent mouse actions on edges and corners; keep the global menu or taskbar flush with the display wall so primary targets gain infinite depth.

Forms and product CTAs

Put Submit / Save near the last field the user edits; enlarge the full clickable label+icon region, not only the glyph, so W matches what users aim at.

Touch and mobile layouts

Grow hit areas and reduce travel between sequential taps; do not assume screen edges help—fingers can overshoot past glass, so edge placement may hurt more than help.

Physical tools and home controls

Make emergency or daily switches large and near the natural hand path (light switches by the door, stove knobs within easy reach); treat tiny distant toggles as high-ID hazards for kids and older adults alike.

​
Case Study

In 1978, Stuart K. Card, William K. English, and Betty J. Burr applied Fitts-style analysis to text selection on a CRT, comparing a mouse, a rate-controlled isometric joystick, step keys, and text keys (Ergonomics, Vol. 21, No. 8, pp. 601–613). Movement time tracked index of difficulty across devices; the mouse showed the highest pointing performance—commonly cited around 10.4 bits/s index of performance in that text-selection setting, in the same ballpark as Fitts’s original motor estimates. The mouse beat the joystick and key-based alternatives on the measured pointing task. That evidence helped justify commercial pointing devices: later accounts of Card’s work at Xerox PARC note the evaluation as a major factor in the mouse’s commercial introduction. The boundary note is important: laboratory D–W conditions and adult expert users do not automatically transfer to touchscreens, accessibility constraints, or tasks where searching and deciding dominate pure pointing.

​
Boundaries and Failure Modes

Boundary 1: Touch and unconstrained space
Screen-edge “infinite width” assumes a hard stop for the pointer. On touch, the finger can leave the display; edge targets can become harder, not easier. Boundary 2: Not only size theater
If users cannot see or trust the hit area (tiny glyph, invisible padding only), they still slow down in the corrective phase. Cognitive search, choice overload, and clutter can dominate motor ID. Common misuse: Treating Fitts as “always make everything big.” Crowding large targets raises wrong-target errors; safety-critical reject actions may need higher difficulty (smaller, farther, or gated) so they are hard to hit by accident—the opposite of primary CTAs.

​
Common Misconceptions

These traps mix motor difficulty with aesthetics, device physics, or decision load.
No. Fitts modeled human aimed movement; limbs, tools, and many pointing devices have been studied. Screens are a major application, not the exclusive domain.
No. The classic relationship is logarithmic in the difficulty index: longer travels add time, but not as a simple linear multiple of distance alone.
No. Pointing ease can fight layout density, mis-tap risk, and visual hierarchy. Optimize high-frequency or high-cost actions first; leave rare destructive actions harder to acquire.
Nearby laws separate pointing cost, choice cost, and system fragility.

Hick's Law

How option count slows choice reaction—pair with Fitts when menus both require deciding and pointing.

Weber–Fechner Law

Ratio-based compression in perception; Fitts uses a related logarithmic framing for motor difficulty.

Yerkes–Dodson Law

Arousal and performance; stress can amplify misses on high-ID targets.

Murphy's Law

If a tiny control can be mistapped under pressure, plan as if it will be.

Gall's Law

Simple working controls beat complex layouts that scatter high-ID targets.

Choice Overload

Too many options tax preference and regret—beyond pure pointing time.

​
One-Line Takeaway

For every frequent action, cut the distance, grow the target—or put the pointer where overshoot is impossible—then measure misses, not only opinions.