Drive Venture research guide
Why supervised practice matters.
Experience matters. Variety matters. Supervision matters. But the evidence does not prove that exactly 50 hours is a universal safety threshold.
Policy benchmark
Why do states use 50 hours?
Fifty hours is widely used in U.S. graduated driver licensing, or GDL. In Drive Venture’s current review of all 50 states and Washington, D.C., 29 jurisdictions use exactly 50 supervised hours in the ordinary teen pathway we identified. Other states require different totals, allow driver-education alternatives, or use age- and pathway-specific rules.
The benchmark also appears in the National Novice Teen Driver Education and Training Administrative Standards, which recommended at least 50 hours of supervised practice in a wide variety of increasingly challenging circumstances. That is a policy and training standard—not proof that hour 50 causes a specific reduction in crashes.
50 hours is widely used
Many states cluster around 40–50 hours, often with a nighttime component. Requirements still vary significantly.
Best-practice targets differ
IIHS currently describes at least 70 supervised hours as a best-practice GDL component. Other countries recommend still more.
No magic cutoff
U.S. component studies have not established a particular supervised-hours calibration as the cause of a specific fatal-crash reduction.
What the data show
The strongest recent U.S. evidence supports more practice—with important limits.
A 2025 National Academies report, sponsored through the Behavioral Traffic Safety Cooperative Research Program by GHSA and NHTSA, analyzed naturalistic driving data collected from instrumented vehicles. In the supervised-practice dataset, the lower-exposure group averaged 17 hours of practice per teen, while the higher-exposure group averaged 52.7 hours.
Teens in the higher-exposure group later had lower rates of safety-critical events and kinematic risky-driving events than teens in the lower-exposure group. Virginia Tech summarized the crash and near-crash rates as approximately 12 events per 1,000 independent-driving hours for the higher-exposure group versus 17 for the lower-exposure group.
| Evidence | Supports | Does not prove |
|---|---|---|
| 2025 U.S. naturalistic study | More supervised exposure was associated with fewer later safety-critical and risky-driving events. | That 50 is an exact causal threshold, or that hours alone explain the difference. |
| U.S. GDL evaluations | Comprehensive staged licensing reduces teen crashes and fatalities. | That all benefits come from supervised-hour requirements. |
| Early novice-driving research | Crash risk declines substantially as months of independent experience accumulate. | That the riskiest solo miles can be converted directly into an equivalent number of supervised hours. |
| International studies | Longer supervised-learning periods can increase pre-license experience and may reduce later crash risk. | One universal hour requirement that transfers unchanged across licensing systems and countries. |
Learning through exposure
What does driving experience build?
A 2026 international literature review concluded that accumulated experience is most consistently associated with better hazard perception and management of attention. The review also identified improvements in vehicle handling, understanding personal limits, speed control, and road positioning.
That provides a reasonable explanation for why supervised practice is valuable: learners can encounter ordinary traffic situations while an experienced adult helps them anticipate hazards and make decisions. However, the research does not establish rigid stages such as “vehicle handling is learned in hours 1–20” and “hazard perception begins in hours 20–50.” Learning rates differ by driver and by the quality of practice.
Vehicle control
Repeated practice can make basic steering, braking, scanning, and positioning less effortful.
Hazard perception
Exposure to real traffic helps learners recognize early cues and anticipate developing conflicts.
Attention management
Experience can improve how drivers divide attention among the road, mirrors, controls, and other users.
Self-awareness
Guided reflection can help learners better understand their limitations and when conditions exceed their skill.
Quality, not just quantity
Why varied practice matters.
Hours are easy to count; experience is harder to measure. A family can accumulate time on the same familiar route without covering the range of situations a new driver will eventually face. NHTSA’s administrative standard therefore pairs its hour recommendation with practice in a “wide variety of increasingly challenging circumstances.”
The 2025 National Academies report similarly examined road type, time of day, route familiarity, passengers, speeding, and risky-driving events. Its broader lesson is that practice exposure is multidimensional. The best target is not merely a filled clock—it is broader, appropriately supervised experience.
- Lighting: daylight, dusk, and darkness
- Roads: neighborhood, rural, multilane, and limited-access
- Traffic: light traffic before more complex congestion
- Weather: rain, wind, glare, and—where locally appropriate—winter conditions
- Tasks: turns, merging, lane changes, parking, navigation, and hazard scanning
- Familiarity: known routes first, then unfamiliar routes with active coaching
Practice should progress only when the learner and supervising adult judge the conditions appropriate. Families should not seek dangerous weather or traffic solely to complete a category.
International context
International evidence strengthens the case for more experience, not one exact number.
Sweden lowered the minimum age for supervised learning from 17½ to 16, creating a much longer opportunity to practice before licensure. Evaluations found substantially more pre-license practice and lower post-license crash risk among participants who used the longer learning period. Because this was a policy evaluation rather than a randomized trial, the result supports—but does not prove—a causal benefit from a particular number of hours.
A recent Norwegian literature review reached a similarly careful conclusion: motivating more practice under safer supervision is sensible, but the authors explicitly declined to endorse one specific hour threshold. They also emphasized that experience varies in quality and diversity.
Evidence guardrails
Claims Drive Venture will not make.
“Exactly 50 hours reduces fatal crashes by X%.”
Unsupported. GDL evaluations measure packages of rules, and research has not isolated exactly 50 hours as a universal causal threshold.
“Fifty hours equals 1,500–2,000 miles.”
Not reliable. Mileage varies dramatically with speed, traffic, route, and lesson structure.
“Ten nighttime hours is the scientifically established minimum.”
Misleading. Ten hours is common in state rules, but research has not established it as a universal threshold.
“The first 50 supervised hours replace the riskiest 50 solo hours.”
Unsupported. Supervised and independent driving differ in exposure, feedback, responsibility, and risk-taking.
For teens and grownups
Practical guidance for better supervised practice.
- Meet your state requirement.Confirm the current rule with the official licensing agency; age and driver-education pathways may change the requirement.
- Do not stop learning at the minimum.An hour requirement is a licensing floor, not a declaration that a learner is ready for every road condition.
- Build complexity gradually.Begin in manageable environments and add traffic, speed, navigation, darkness, and unfamiliar roads as skills improve.
- Track conditions as well as time.A logbook can reveal missing experience—for example, many hours in daylight but little practice at dusk or on multilane roads.
- Coach, then reflect.Before a drive, agree on one or two goals. Afterward, discuss what went well, what surprised the learner, and what to practice next.
- Delay independent driving if needed.Completing the legal minimum does not require a family to conclude that the teen is ready to drive alone.
Sources and methods
Research sources.
Drive Venture prioritizes U.S. government, National Academies, peer-reviewed, and established traffic-safety sources. International work is included when it directly examines accumulated experience or supervised practice. Association is not presented as causation, and evidence for GDL as a whole is not presented as proof for exactly 50 hours.
- National Academies of Sciences, Engineering, and Medicine. Outcomes of Variability in Teen Driving Experience and Exposure: Evidence from Naturalistic Driving Studies. 2025. Sponsored by GHSA and NHTSA.
- Virginia Tech Transportation Institute. “New study: Teen drivers safer with more practice.” 2025. Plain-language summary of the National Academies report.
- National Highway Traffic Safety Administration. Teen Safe Driving and Novice Teen Driver Education and Training Administrative Standards.
- Insurance Institute for Highway Safety. Teenagers and Graduated licensing laws.
- Walshe EA, et al. Licensing Examination and Crash Outcomes Postlicensure in Young Drivers. JAMA Network Open. 2022.
- Masten SV, Foss RD, Marshall SW. Graduated driver licensing program component calibrations and their association with fatal crash involvement. Accident Analysis & Prevention. 2013.
- Russell KF, et al. Graduated driver licensing for reducing motor vehicle crashes among young drivers. Cochrane systematic review.
- Simons-Morton BG, Ehsani JP. Learning to Drive Safely: Reasonable Expectations and Future Directions for the Learner Period. Safety. 2016.
- Gregersen NP, et al. Sixteen years age limit for learner drivers in Sweden—an evaluation of safety effects. Accident Analysis & Prevention. 2000.
- Nævestad T-O, et al. Why does novice drivers’ risk decrease by almost 50% during the first year, and what do they learn? Frontiers in Sustainable Cities. 2026.