Informational article, not legal advice. This is general education about Romanian criminal procedure and public statistics. Aggregated numbers describe a system trend; they do not predict outcomes in an individual case, and this article does not promise results.

0) How to use this article

• 2 minutes: take the Quick Quiz and write down your score.
• 2 minutes: read the Funnel section (the key takeaway, in numbers).
• 1 minute: tick the checklist (what matters most in practice: timeline + evidence).

If you only want the main idea, jump to “4) Funnel: out of 100 cases, where do they end up”.

What sources I use (and their limits)

Before any number: I only use figures that can be traced to an official, publicly accessible document, with a clear page/section reference.

• Public Ministry (Romania – Prosecutor’s Office): the official annual activity report for 2024. In this article I use the statistical annex “Documentary regarding the activity of the Public Ministry in 2024”, p. 71 (total “cases solved on the merits”) and p. 72 (breakdown: indictment vs non-indictment; dismissal; waiver of prosecution). Official access page: Public Ministry – Activity report.
• Romanian Police: the public 2024 activity report page, which includes aggregate indicators (cases handled together with prosecution units; cases solved; proposals to send cases to court). Official access page: Romanian Police – Activity report 2024.
• Romanian Criminal Procedure Code (CPP): legal meaning of the procedural outcomes. Official consolidated text: Criminal Procedure Code (legislatie.just.ro) – art. 315 (dismissal / “clasare”), art. 318 (waiver of prosecution / “renunțare la urmărire penală”), art. 478+ (plea agreement / “acord de recunoaștere”).

Limits I respect (no “creative math”): (1) Romanian Police “cases/files” and Public Ministry “cases solved on the merits” are different reporting units and methodologies, so I do not merge them into one single funnel. (2) In the Public Ministry table used here, “sent to court” is reported as indictments + plea agreements together; therefore, I cannot confirm the number of plea agreements separately from indictments from this same table.

Useful reminder: aggregated statistics are not a prediction for your specific file.

1) Simple dictionary (cards)

Recap: dismissal and waiver are non-indictment outcomes; indictments and plea agreements are “sending the case to court” outcomes (but in the Public Ministry table used here, indictments + plea agreements are reported together).

2) Quick quiz (2 minutes): 7 questions + score

Pick A/B/C. Write your answers down, then check the key and your score.

1. “Dismissal / clasare” is best described as:
   A) a quick conviction
   B) a non-indictment procedural outcome under the law
   C) a postponed trial
2. Waiver of prosecution means “nothing happened”:
   A) always true
   B) false; it is a lawful outcome with statutory criteria (incl. public interest)
   C) true only if the victim agrees
3. Sending a case to court automatically means conviction:
   A) yes
   B) no
   C) only for serious offences
4. In public statistics, “a case file” and “a person” are always the same unit:
   A) yes
   B) no
   C) it depends on the county
5. A screenshot is always enough evidence:
   A) yes, if it looks clear
   B) no; context and authenticity can be challenged
   C) yes, if it has a date visible
6. A plea agreement bypasses the court:
   A) yes
   B) no; the court verifies and decides under the CPP
   C) sometimes
7. The most reliable “practical accelerator” is:
   A) a clean timeline and organized evidence
   B) a very long complaint
   C) posting online

Answer key + scoring

• Correct answers: 1B, 2B, 3B, 4B, 5B, 6B, 7A.
• Score: 0–2: recap needed; 3–5: solid; 6–7: excellent.

Reminder: even a perfect quiz score does not turn statistics into a prediction for an individual file.

3) The raw numbers (table + notes)

These are the indicators used later in the funnel. Note the two reporting worlds: Public Ministry (cases solved on the merits) and Romanian Police (files handled/solved together with prosecution units).

Indicator	Value + source
Public Ministry – cases solved on the merits (total)	541,671 (Public Ministry 2024 activity report, “Documentary…”, p. 71)
Public Ministry – sent to court (indictments + plea agreements, reported together)	51,111 (Public Ministry 2024 activity report, “Documentary…”, p. 72)
Public Ministry – non-indictment (total)	490,560 (Public Ministry 2024 activity report, “Documentary…”, p. 72)
↳ of which: waiver of prosecution	60,153 (Public Ministry 2024 activity report, “Documentary…”, p. 72)
↳ of which: dismissal (“clasare”)	430,407 (Public Ministry 2024 activity report, “Documentary…”, p. 72)
Romanian Police – files handled (together with prosecution units)	over 1.7 million (Romanian Police 2024 activity report page, p. 1 of the PDF excerpt)
Romanian Police – files solved	502,584 (Romanian Police 2024 activity report page, p. 1 of the PDF excerpt)
↳ of which: proposals to send the case to court	50,064 (Romanian Police 2024 activity report page, p. 1 of the PDF excerpt)

What I do NOT compare (to avoid misleading you): (1) Police “proposal to send to court” is not the same as the prosecutor’s “sending the case to court”. (2) Police “files” are not the same reporting unit as Public Ministry “cases solved on the merits”. (3) In the Public Ministry table used here, “sent to court” is an aggregate of indictments + plea agreements, so I do not split plea agreements from indictments without an official split in the same source.

4) Funnel: “Out of 100 cases, where do they end up?”

This funnel is built exclusively on the Public Ministry data for cases solved on the merits (2024). I do not mix Police numbers into this funnel.

(A) Simple version (percentages + formulas)

1. Base: total cases solved on the merits = 541,671 (100%).
2. Sent to court (indictments + plea agreements, aggregate): 51,111 / 541,671 = 9.4%. Here X = 51,111 “sent to court”; Y = 541,671 total solved on the merits.
3. Non-indictment (total): 490,560 / 541,671 = 90.6%. Here X = 490,560 non-indictment; Y = 541,671 total solved on the merits.
4. Waiver of prosecution (out of total solved on the merits): 60,153 / 541,671 = 11.1%.
5. Dismissal / “clasare” (out of total solved on the merits): 430,407 / 541,671 = 79.5%.

If your question is “within non-indictment outcomes, what share is waiver vs dismissal?” then the base changes (different question, different denominator): waiver = 60,153 / 490,560 = 12.3%; dismissal = 430,407 / 490,560 = 87.7%.

(B) Visual version (ASCII/emoji blocks) + legend

Base: 541,671 cases solved on the merits (Public Ministry, 2024)

[🟩🟩🟩🟩🟩🟩🟩🟩🟩🟩] 100% = total solved on the merits

[🟦🟦] 9.4% = sent to court (indictments + plea agreements, aggregate) = 51,111 / 541,671
[🟥🟥🟥🟥🟥🟥🟥🟥🟥] 90.6% = non-indictment (total) = 490,560 / 541,671

  [🟨🟨] 11.1% = waiver of prosecution (out of total solved on the merits) = 60,153 / 541,671
  [⬛⬛⬛⬛⬛⬛⬛⬛] 79.5% = dismissal / “clasare” (out of total solved on the merits) = 430,407 / 541,671

Legend:
🟩 base (100%)
🟦 sent to court (aggregate: indictments + plea agreements)
🟥 non-indictment (total)
🟨 waiver of prosecution
⬛ dismissal / “clasare”

Recap: in this official dataset (Public Ministry, 2024), most “solved on the merits” outcomes are non-indictment, while “sent to court” is a smaller subset. That does not mean an individual file is “doomed” or “guaranteed” – it means evidence, legal classification, and procedural specifics matter.

5) Choose your scenario (interactive decision tree)

Pick the closest scenario. Each one has three branches: strong evidence / weak evidence / unknown offender. This is general procedural logic, not individualized advice.

1) Minor theft / neighbor conflict

If you have strong evidence

Clear timeline, witnesses, CCTV, identifiable offender. In practice, verifiable evidence reduces ambiguity and helps authorities test the facts.

If evidence is weak

When the file becomes “your word vs their word” without external confirmation, non-indictment outcomes (including dismissal) become more likely. You help by pointing to specific checks (time, place, cameras, witnesses).

If the offender is unknown

Identification becomes the gate. Without realistic leads to identify the offender, the evidentiary threshold may not be met.

2) Assault / threats

If you have strong evidence

Medical documents, consistent witness accounts, complete message history, recordings (within legal limits). Corroboration is key.

If evidence is weak

Authorities look for external confirmation. A clean timeline and consistent statements matter a lot.

If the offender is unknown

Less common, but possible in street incidents. Early identification (cameras, witnesses, location data) can be decisive.

3) Fraud / scam

If you have strong evidence

Contracts, transfers, complete conversations (not fragments), identity indicators, and a clear money trail. The practical goal is a demonstrable chain: promise → payment → damage.

If evidence is weak

Isolated screenshots are easy to challenge. Without payment traceability and documents, the criminal-law proof threshold can be hard to reach.

If the offender is unknown

Online identification often depends on technical traces and the money trail. Preserve IDs, links, email headers, transaction references, and full chat exports.

4) Traffic accident

If you have strong evidence

Official minutes, photos, dashcam/CCTV, witnesses, and sometimes expert analysis. Technical reconstruction can change initial impressions.

If evidence is weak

Competing narratives appear. Details like positions, traces, and consistency matter more than emotions.

If the offender is unknown

Common in “hit and run” situations. Early camera checks and immediate witness data are often decisive.

5) Online report + digital evidence

If you have strong evidence

Full conversation exports, original files, links, IDs, timestamps, and a separate backup. Strong digital evidence is complete and verifiable.

If evidence is weak

Decoupled screenshots are fragile. Preserve sources: exports, originals, and context, not just fragments.

If the offender is unknown

Linking an account to a person can be complex. Early preservation of account IDs, transaction details, and platform data matters.

Recap: in every scenario, verifiable evidence + coherent timeline reduce uncertainty. And uncertainty is where non-indictment outcomes often live.

6) Myth vs Fact (at least 10)

Myth: “If a case is dismissed, it proves I’m innocent.”
Fact: Dismissal is a procedural non-indictment outcome with legal grounds. See CPP, art. 315.

Myth: “Waiver of prosecution is mercy without rules.”
Fact: Waiver has statutory conditions and criteria (including public interest). See CPP, art. 318.

Myth: “If I have one witness, I will win.”
Fact: A witness is one piece of evidence; credibility, consistency, and corroboration matter.

Myth: “A screenshot is always enough.”
Fact: Screenshots can be challenged; originals, context, and authenticity checks matter.

Myth: “If the prosecutor sends the case to court, conviction is guaranteed.”
Fact: The court decides after its own assessment and trial.

Myth: “A plea agreement bypasses the court.”
Fact: The court verifies and decides under the law. See CPP, art. 478 et seq.

Myth: “If I withdraw my complaint, the file closes automatically.”
Fact: It depends on the offence type and legal regime; there is no universal rule.

Myth: “If the police propose sending to court, the prosecutor must follow.”
Fact: A police proposal is not the final decision; the prosecutor decides procedural outcomes.

Myth: “If I’m not called quickly, it’s a cover-up.”
Fact: Delay alone proves nothing; workload, procedural steps, and priorities exist.

Myth: “The truth will emerge by itself.”
Fact: Evidence must be gathered, preserved, and presented coherently. Without verifiable evidence, facts may remain unconfirmed.

7) One-minute checklist (tickable)

Tick what you already have. The empty boxes are your realistic to-do list (no promises, just practical logic).

• ☐ I have a clear timeline (dates/times/places, in order).
• ☐ I have witnesses identified (names + how they can be reached).
• ☐ I have relevant documents (contracts, receipts, bank statements, official papers).
• ☐ I preserved digital evidence completely (full exports, originals, links, IDs – not only screenshots).
• ☐ I have a separate backup of evidence (cloud or external drive).
• ☐ My report is coherent (facts first, interpretations later).
• ☐ I track communications and deadlines (file number, responses, dates).
• ☐ I avoid contradictions (same facts, same order, same wording).

Recap: coherence + verifiable evidence are the “language” of criminal procedure. Statistics are not a substitute for evidence.

8) Plea agreements (mini-simulation + Stop & Think)

This is a simplified simulation to explain the mechanism (informational only, not a recommendation for any specific file).

1. The prosecutor evaluates facts and legal classification.
2. The defendant (with counsel) considers admitting the offence under legal conditions.
3. The terms are set within statutory limits (it is not “anything goes”).
4. The agreement is submitted to the court.
5. The court verifies legality and conditions under the CPP.
6. The court admits or rejects; there is no automatic guaranteed outcome.

Stop & Think (3 quick questions)

1. What does the court verify? Legal conditions and the framework in the CPP (art. 478 et seq.).
2. What does “admitting” mean? A procedural position with serious consequences; it must be fully understood.
3. What does a plea agreement NOT do? It does not bypass the court and it does not guarantee a “convenient” outcome.

Legal source: CPP, art. 478 et seq.

9) FAQ (short questions, short answers) – minimum 12

In aggregate, what’s more frequent: sending to court or non-indictment?

In the Public Ministry 2024 dataset for “cases solved on the merits”, non-indictment is the majority: 490,560 / 541,671 = 90.6% (Public Ministry 2024 activity report, “Documentary…”, p. 71–72).

How many cases reach court in this dataset?

“Sent to court” is reported as indictments + plea agreements together: 51,111 / 541,671 = 9.4% (Public Ministry 2024 activity report, “Documentary…”, p. 71–72).

Does dismissal (“clasare”) mean I’m innocent?

Not automatically. It is a procedural non-indictment outcome with legal grounds. See CPP, art. 315.

Is waiver of prosecution basically “forgiveness”?

No. It is a lawful outcome with statutory criteria (including public interest). See CPP, art. 318.

Can we see plea agreements separately from indictments in this Public Ministry table?

No. In the table used for the funnel, “sent to court” is an aggregate (indictments + plea agreements). I cannot confirm a separate number for plea agreements from that same table.

Within non-indictment outcomes, which is more common: dismissal or waiver?

Within non-indictment as the denominator: waiver = 60,153 / 490,560 = 12.3%; dismissal = 430,407 / 490,560 = 87.7% (Public Ministry 2024 activity report, “Documentary…”, p. 72).

If the police propose sending to court, is it guaranteed?

No. A police proposal is not the prosecutor’s final procedural decision.

Why do Police numbers look different from Public Ministry numbers?

Different reporting units and methodologies (“files handled/solved” vs “cases solved on the merits”). That’s why this article does not merge them into one funnel.

Can I estimate my chances from the funnel?

No, not responsibly. The funnel is aggregated statistics; your specific case depends on facts, evidence, and legal classification.

What helps most in practice?

A clear timeline and verifiable evidence organized in a way that can be checked. Start with the checklist above.

Does withdrawing a complaint always stop the case?

No. It depends on the offence type and legal regime; there is no universal rule.

I received a dismissal decision. What now?

There are legal mechanisms to challenge certain decisions, but the concrete steps depend on the file and documents. This article is informational, not individualized legal advice.

10) Mini-timeline (retention bonus): 7 typical stages

1. Report: complaint/denunciation/ex officio initiation.
2. Registration & allocation: competence and assignment.
3. Evidence gathering: statements, documents, technical checks, expert reports, digital data.
4. Offender identification: often the practical “gate”.
5. Evidentiary assessment: can the evidence meet the required threshold?
6. Prosecutor’s outcome: non-indictment (dismissal/waiver) or sent to court (indictment/plea agreement), as applicable.
7. If it reaches court: judicial review/trial and final decision.

Recap: not every case follows the same speed or path, but the logic repeats: evidence → legal framing → procedural outcome.

11) Conclusion: 3 things to remember

• Aggregates do not predict your file: statistics describe the system, not your case.
• Procedural outcomes have legal meaning: dismissal/waiver are legal mechanisms, not moral labels.
• Verifiable evidence beats intensity: complete, consistent, easy to check.

12) Sources

• Public Ministry – Activity report (official access page) – 2024 report, statistical annex “Documentary regarding the activity of the Public Ministry in 2024”, p. 71–72 (figures used for the funnel: total solved on the merits; indictment vs non-indictment; dismissal; waiver of prosecution).
• Romanian Police – Activity report 2024 (official access page) – aggregate indicators shown on the report page / PDF excerpt (files handled together with prosecution units; files solved; proposals to send to court).
• Criminal Procedure Code (legislatie.just.ro) – art. 315 (dismissal / “clasare”), art. 318 (waiver of prosecution), art. 478 et seq. (plea agreement).