CSA Scores Explained
CSA Scores Explained: what they are, how they’re calculated, and how to improve them.
CSA (Compliance, Safety, Accountability) is FMCSA’s safety measurement framework that organizes roadside inspection and crash data into categories (BASICs). Carriers get percentile rankings that can influence audits, insurance pricing, and shipper/broker confidence. This guide breaks down the parts, what matters most, and practical steps to reduce risk.
- Understand each BASIC and which violations move the needle the fastest.
- Interpret percentiles (and why they’re not “your score” in isolation).
- Build a repeatable process: DVIR discipline, maintenance docs, and driver coaching.
- Reduce the “surprise factor” before inspections, brokers, and compliance reviews.
CSA Scores Explained: What You’re Seeing, What It Means, and How to Fix It
What people call a “CSA score” is usually a BASIC percentile in FMCSA’s Safety Measurement System (SMS). It’s a relative ranking against similar carriers (0–100), where higher percentiles mean worse safety performance.
This briefing explains CSA/SMS in plain language, shows how percentiles are built (time weights + severity weights), and gives you an improvement plan that actually moves the numbers. Educational content only (not legal advice).
Related (internal): DOT Audit Guide • HOS / ELD Simulator • Clearinghouse • Insurance
Three layers people confuse (and why it matters)
Most bad decisions happen because carriers mix up CSA, SMS percentiles, and safety ratings. They are not the same thing.
FMCSA’s safety enforcement and compliance program. It’s the umbrella that includes SMS as a prioritization tool.
Operational performance percentiles by BASIC (0–100). Higher percentile = worse relative performance.
A formal safety fitness determination generally issued after an onsite investigation. It’s not your BASIC percentile.
Straight truth
- BASIC percentiles are used to prioritize carriers for interventions (e.g., warning letters, investigations).
- Percentiles do not automatically change your official safety rating.
- FMCSA’s public SMS display includes a disclaimer that the SMS data is not a safety rating.
Quiet winners don’t argue about “scores.” They remove repeat violations and improve documentation until the percentiles drop.
Driver misconception (fast)
CSA does not give FMCSA the authority to revoke a CDL. CDL licensing action comes from state CDL agencies, not CSA. For driver-level screening, carriers typically use PSP (5-year crash / 3-year inspection history).
Driver note: a PSP record is a safety record and does not provide a “score.”
As-of: SMS Methodology dated September 2025; eCFR and CSA pages should be checked any time you’re making a high-stakes decision.
The 7 BASICs (and what’s public)
SMS organizes safety data into seven BASICs. Public SMS shows five BASICs. Two BASICs are typically visible only to the carrier when logged in.
Behaviors like speeding, reckless driving, texting/handheld phone, lane violations.
Log compliance, fatigue-management rules, and related HOS violations.
Mechanical condition, inspections, defects, and OOS-related maintenance violations.
Driver qualification issues, medical, licensing, and qualification documentation.
Drug/alcohol-related inspection violations; separate from Clearinghouse reporting.
Crash involvement patterns based on state-reported crashes; crashes reviewed “Not Preventable” can be excluded from the measure.
Hazmat compliance BASIC; only relevant to HM carriers and typically not public on SMS display.
Tip: if you’re working on compliance readiness, pair this page with DOT Audit Guide and keep evidence packs per BASIC.
How SMS math works (why recent violations hurt more)
Percentiles aren’t just “counts.” SMS applies severity weights and time weights, and compares carriers to peers with similar event counts.
Each violation has a severity weight (higher = worse). Some OOS conditions add additional weight in applicable BASICs.
Past 6 months = weight 3; 6–12 months = weight 2; 12–24 months = weight 1. Older drops off.
Within a BASIC, the sum of severity weights per inspection is capped (then time-weighted), then carriers are ranked into a 0–100 percentile.
What a percentile really means
- 0 ≈ best performance among peers (not “zero violations forever”).
- 100 ≈ worst performance among peers.
- Higher = worse relative safety performance.
- Comparisons are made within BASIC groupings based on event counts (crashes/inspections/violations).
In practice: one recent high-severity violation can matter more than several older low-severity issues.
The cleanest fix strategy
Don’t chase every point. Chase the top two: repeat violations and recent violations. Build a weekly review habit that prevents repeats and documents corrective actions.
If your percentiles are “stuck,” you usually have a process problem (training, inspections, maintenance cadence), not a “math problem.”
If your business is driver-heavy, also remember the driver view: PSP shows a driver’s 5-year crash and 3-year inspection history.
Thresholds & “alert” status (what triggers attention)
Carriers may be prioritized for interventions when BASIC percentiles are at or above established Intervention Thresholds and/or when Acute/Critical violations were found in investigations.
Illustrative only: this is an example of how percentiles can be interpreted versus a threshold line (not your real SMS).
Reminder: a BASIC “alert” symbol is a prioritization indicator and is not intended to imply a federal safety rating.
Fixing bad data (DataQs) — how corrections actually happen
If data is incomplete or incorrect, carriers and drivers can submit a Request for Data Review (RDR) through DataQs. This is also how adjudicated citation outcomes can affect how violations appear in SMS.
Confirm the inspection/crash belongs to your USDOT, unit, driver, and date/location.
Incorrect data? Missing correction? Adjudicated citation outcome?
If a citation was changed/dismissed, include certified court documentation and inspection details.
Monitor status and respond if the state/FMCSA requests more info.
Even if corrected, fix root cause so it doesn’t repeat next week.
Adjudicated citations: practical impact
- Dismissed / not guilty: violation can be removed from SMS.
- Convicted of different charge: severity weight can be set to 1 (and may not be subject to OOS weight).
- How it happens: you submit certified judicial documentation via DataQs to initiate the process.
Do not assume a “dismissed ticket” automatically fixes SMS. The workflow is usually “court → certified docs → DataQs → update.”
Fastest wins
Data correction helps, but the biggest percentile drops come from stopping repeat violations. Fix the top two repeat items and your newest 6-month window improves first (time weight = 3).
Build a weekly review routine: inspection notes, driver coaching, maintenance defect loop, and proof.
Related: if your risk includes drug/alcohol program issues, read Clearinghouse Explained.
Fix-it playbook (what moves percentiles)
If you want percentiles to drop, you need fewer repeat violations in the newest 6 months. This is the “quiet winner” approach: fix process + prove it.
The weekly close (CSA version)
- Review all inspections (even “clean” ones) and note what saved you.
- Count repeats by BASIC and by driver/unit.
- Coach + document corrective action (dated).
- Maintenance loop: defect in → repair out → proof filed.
- HOS loop: errors in → training out → review logs again next week.
You win CSA by being boring: consistent checklists, consistent training, consistent proof.
BASIC-by-BASIC focus (simple)
Don’t try to fix seven categories at once. Fix the ones that are (a) above threshold, and (b) repeating. Your fastest ROI is often: HOS (process), Maintenance (cadence), and Unsafe Driving (behavior).
Tie each fix to evidence: training log, signed checklist, work order, review sheet.
Quick internal link pack
- DOT Audit Guide — build the binder + proof habits.
- HOS / ELD Simulator — tighten log discipline.
- Truck Insurance Basics — why claims + safety culture change premiums.
- Carrier Onboarding — standardize paperwork and ops.
30 / 60 / 90 day plan (to materially improve CSA/SMS)
The fastest measurable improvement happens by shrinking repeat violations in the newest 6 months (time weight = 3), and by building proof that your safety system is real.
Pick 2 BASICs. Fix the top repeat violation with training + checklist + supervisor review.
Weekly HOS review + maintenance defect loop. Document corrective action after every inspection.
Turn the fixes into SOP: onboarding, pre-trip, log review, PM cadence, coaching cadence.
Keep an evidence pack per BASIC. If the agency asks, you can prove your system quickly.
If you want a measurable KPI: track “repeat violations per 10 inspections” by BASIC and push that down each week.
FAQ
What is a “CSA score” really?
Do BASIC percentiles change my official safety rating?
How long do violations affect SMS?
Why are some BASICs not public?
How do I correct incorrect inspections/crashes?
Does CSA revoke a CDL?
Quick next steps
If you’re trying to move percentiles (not argue about them), this is the cleanest sequence.
- Step 1: Identify the top two BASICs that are (a) highest, and (b) repeating.
- Step 2: Fix the repeat driver/unit pattern with one checklist + one review owner.
- Step 3: Document every corrective action (dated) and file proof in an evidence pack.
- Step 4: Run a 15-minute “weekly close” until repeats fall in the newest 6-month window.
Evidence pack (simple)
- HOS: weekly log review notes + training acknowledgements.
- Maintenance: DVIR/defect reports + work orders + PM schedule adherence.
- Unsafe driving: coaching notes + policy refresh + any telematics events review.
- Driver fitness: DQ file completeness checks (med cards, CDL status, annual review).
Educational content only (not legal advice).
Back to top
Related: DOT Audit Guide • Clearinghouse • Insurance • Driver Job Market