In this article:
- What a driver scorecard is, and what it isn't
- The metrics: what to measure and where it comes from
- Why some metrics should weigh more than others
- How to normalize, so the scorecard is fair
- A worked example: two drivers, one quarter
- How often to score, and who sees it
- How to use the scorecard without losing drivers' trust
- Common driver scorecard mistakes
- About Off Duty Not Driving Rewards
A driver scorecard is a regular, per-driver summary of safety performance built from records the fleet already keeps. Good ones score crashes, roadside inspection results, hours-of-service violations, moving violations, and inspection paperwork, normalized for miles or inspections, and weighted toward the behaviors FMCSA links most closely to crash risk. Then they're used for coaching and recognition, consistently.
The rest of this guide covers the metrics and where each comes from, how to weight and normalize them, a worked example with two drivers, and how to use the result without losing drivers' trust.
Off Duty Not Driving Rewards is a rewards program for trucking fleets. Contact our team to learn more.
What a driver scorecard is, and what it isn't
A driver scorecard is a per-driver view. The CSA Safety Measurement System (SMS) is a carrier view: it ranks your company against peers in seven categories called BASICs, using roadside inspections and crashes from the past two years (FMCSA CSA: Measure). A scorecard breaks the same kind of data down by driver so you can see who needs coaching and who has earned recognition.
It is also not a legal record by itself. The records behind it are, though: your accident register, driver qualification files, records of duty status and inspection reports all have federal retention rules. Build the scorecard on top of them, not instead of them.
The metrics: what to measure and where it comes from
Each metric below is tied to a record the regulations already make you keep, or to data FMCSA itself scores.
| Metric | What counts | Where the data comes from | Rule or source |
|---|---|---|---|
| Preventable crashes | DOT-recordable crashes: a death, an injury treated away from the scene, or a tow for disabling damage | Accident register (kept 3 years) | 49 CFR 390.5, 390.15 |
| Unsafe driving violations | Roadside violations for speeding, texting, hand-held phone use, reckless driving and similar | Roadside inspection reports | Unsafe Driving BASIC (CSA: Measure) |
| Moving-violation convictions | Speeding, reckless driving and other convictions on the driver's record | Motor vehicle record (MVR), pulled at least every 12 months | 49 CFR 391.25 |
| Hours-of-service violations | Driving past 11 hours, past the 14-hour window, missing the 30-minute break, log errors | ELD records and roadside reports | 49 CFR 395.3, 395.8 |
| Clean inspection rate | Roadside inspections with no violation ÷ all roadside inspections | Roadside inspection reports | SMS measures divide by all relevant inspections, clean ones included (SMS Methodology v3.21, pp. 2-5 to 2-6) |
| Out-of-service violations | Any driver or vehicle out-of-service order | Roadside inspection reports | CVSA out-of-service criteria, updated each April 1 (CVSA) |
| Inspection paperwork | Post-trip inspection reports completed, defects listed | Driver vehicle inspection reports (DVIRs) | 49 CFR 396.11 |
| Training and meetings | Required sessions attended | Training records | FMCSA Safety Management Cycle (FMC-CSA-12-019) |
If your trucks have telematics, you can add speeding or hard-braking events. FMCSA's Safety Management Cycle suggests monitoring drivers' speed with "an electronic or manual movement record," such as GPS. Define the event threshold in writing first, and score events per 1,000 miles, not as raw counts.
The Off Duty Brief
Get FMCSA changes to CSA scoring and roadside inspections when they happen.
Plus one practical guide a month. Free, and one click to leave.
Why some metrics should weigh more than others
FMCSA's own methodology says three BASICs, Unsafe Driving, Crash Indicator and Hours-of-Service Compliance, "have the strongest associations to crash risk" (SMS Methodology v3.21, p. 2-8). That's why their intervention thresholds are lower: 65% for most trucking fleets, against 80% for Vehicle Maintenance, Controlled Substances/Alcohol and Driver Fitness (Table 2-2).
A scorecard can follow the same logic. Put the heaviest weight on crashes, unsafe driving and hours of service, and lighter weight on paperwork and training. The SMS also adds extra weight to out-of-service violations and counts recent events more heavily: three times as much in the past 6 months as in the second year (pp. 2-4 to 2-5). A scorecard that rewards recent improvement moves in the same direction as your carrier data.
Driver history matters too. Among drivers of large trucks (over 10,000 lb gross vehicle weight rating) in fatal crashes in 2024, NHTSA found 17.0% had at least one speeding conviction in the prior 5 years, and 19.3% had a previously recorded crash (NHTSA, DOT HS 813 816). Your annual MVR review must give "great weight" to violations like speeding and reckless driving (391.25), so those belong on the scorecard as well.
How to normalize, so the scorecard is fair
Raw counts punish the drivers who drive the most or get inspected the most. Normalize each metric:
- Per million miles for crashes: crashes ÷ miles × 1,000,000. FMCSA uses accident rate per million miles as a factor in safety ratings (49 CFR 385.7).
- Per inspection for roadside results: clean inspections ÷ total inspections. A driver with zero inspections in the period gets a neutral score, not a perfect one or a zero.
- Per 1,000 miles for telematics events.
- Within peer groups: compare regional drivers with regional drivers and long-haul with long-haul. FMCSA does the same thing at carrier level: SMS ranks carriers against peers with a similar number of inspections or safety events (SMS Methodology p. 2-6).
- Only rank drivers with enough data. SMS sets minimums before it assigns a percentile: for example, at least three relevant driver inspections for HOS Compliance and at least five relevant vehicle inspections for Vehicle Maintenance (SMS Methodology pp. 3-12, 3-15). Set your own minimum miles or inspections for new and part-time drivers, and show them their numbers without a rank until they reach it.
A worked example: two drivers, one quarter
Illustration only. This sample scorecard is out of 100 points. Each category starts full and loses points as defined.
| Category (points) | Scoring rule |
|---|---|
| Crashes (30) | −30 for a preventable DOT-recordable crash; −10 for a preventable non-recordable incident |
| Unsafe driving (25) | −10 per roadside unsafe-driving violation or moving-violation conviction |
| Hours of service (15) | −5 per HOS violation; −10 if it led to an out-of-service order |
| Roadside inspections (15) | 15 × (clean inspections ÷ all inspections); a neutral 7.5 if no inspections |
| Paperwork (10) | −2 per missed or incomplete DVIR |
| Training (5) | 5 × (sessions attended ÷ sessions required) |
| Category | Driver A | Driver B |
|---|---|---|
| Crashes | No crash: 30 | One preventable recordable crash: 0 |
| Unsafe driving | One roadside speeding violation: 15 | None: 25 |
| Hours of service | None: 15 | One violation: 10 |
| Roadside inspections | 2 of 3 clean: 10 | 1 of 1 clean: 15 |
| Paperwork | All DVIRs complete: 10 | One missed: 8 |
| Training | 2 of 2: 5 | 1 of 2: 2.5 |
| Total | 85 | 60.5 |
Driver A needs a conversation about speed. Driver B needs a crash review first: was it a driver issue, or a dispatch, route or maintenance issue? FMCSA's Safety Management Cycle tells carriers to "always assess whether an issue is individual or represents a systemic breakdown." For a spreadsheet version of this template, see our driver scorecard template (Excel).
How often to score, and who sees it
- Monthly is a good default. SMS updates monthly, so a monthly scorecard lines up with your carrier data. Weekly works for short feedback loops; quarterly works for rewards.
- Drivers see their own scorecard. A score a driver never sees can't change behavior.
- Keep the history. Your annual driving-record review has to be noted in the driver qualification file (49 CFR 391.51), and a year of scorecards gives that review real content.
How to use the scorecard without losing drivers' trust
FMCSA's Safety Management Cycle sets a standard worth copying: make sure the people monitoring drivers are "applying performance standards fairly, consistently, and equitably; and documenting evaluations" (FMC-CSA-12-019).
In practice:
- Publish the scoring rules before the first scorecard goes out.
- Let drivers dispute data. If a roadside violation was dismissed in court, it shouldn't stay on the scorecard. The same goes for your CSA record, through DataQs. See how to improve your CSA score.
- Coach first. The cycle pairs recognition with immediate feedback and remedial training, and with progressive discipline where needed.
- Recognize good scores visibly. The same FMCSA guidance recommends incentive or recognition programs for clean inspections and no crashes. FMCSA's guidance is a recommendation, not a requirement, and does not endorse any vendor.
- Don't score what drivers can't control, such as delivery windows. FMCSA warns that on-time delivery incentives "could encourage speeding."
Common driver scorecard mistakes
- Scoring raw counts. The highest-mileage driver will always look worst.
- Treating zero inspections as perfect. It means no data, not a clean record.
- Changing the weights mid-period. Drivers stop trusting the score.
- Using the scorecard only for discipline. Then drivers learn to hide incidents rather than avoid them.
About Off Duty Not Driving Rewards
Off Duty Not Driving Rewards is a rewards program for trucking fleets. Contact our team to learn more.
This article is general information, not legal advice. Confirm requirements against the current regulation text and with your compliance counsel.
The Off Duty Brief
FMCSA rule-change alerts plus one practical guide a month, for fleet safety and retention.
Checked against the regulation text before it goes out. One click to unsubscribe.