Forty-eight hours is long enough to feel like you have time, and short enough that you do not. The OSWE exam compresses source review, exploit development, and a technical report into a window where every hour spent on the wrong machine is an hour the report will not get back.
This is what the two days actually look like, and where candidates lose them.
The first four hours set the tone
The instinct is to open the first target and start reading code. The better opening is a map: which machines exist, which are reachable directly, which require a chain, and what the network topology implies about order.
A practical first block:
- Inventory the targets and record what each exposes.
- Pick the one with the clearest attack surface — usually a web application with visible source.
- Map its routes and authentication flow before reading individual functions.
- Write the first note in the report document, so the structure exists before the findings.
Candidates who skip the inventory often spend day one on a machine that was designed to be reached later, through a foothold they have not obtained.
Day one: depth over breadth
The first day is for finishing something. One target taken from source review to working exploit to proof is worth more than three targets half-understood, because the completed one also creates the foothold for the next.
The rhythm that works:
- Read the route, then the handler, then the sink. Do not read the whole codebase.
- Form a hypothesis about the flaw before testing it. The test confirms or kills the hypothesis; it does not discover it.
- Write the exploit as a script immediately. Not after the next target — now, while the session logic is in your head.
- Capture evidence at the moment of success. Screenshot, output log, and the exact command.
By the end of day one, a realistic goal is one target complete and a second mapped.
The night block: the decision candidates get wrong
There is a temptation to work through the night because the clock allows it. The exam is 48 hours, but human accuracy is not. Exploit development at hour thirty with no sleep produces code that fails on a second run and evidence that does not reproduce.
Six hours of sleep is not lost time. It is the difference between a working exploit at hour thirty-four and a debugging session that consumes the report window.
Day two: chains and coverage
The second day is for the objectives that depend on the first, and for the parts of the network that were out of reach. This is where a candidate who took notes on credentials, sessions, and network position moves quickly, and a candidate who did not starts re-enumerating.
Two habits matter here:
- Keep a credential and session inventory. Every account, token, and key, with where it was found and what it reaches.
- Re-read the objective list. It is easy to keep exploring after the required objectives are met, and expensive.
The report window is real work
Most candidates plan to “write it at the end” and discover that a technical report with exploit code takes eight to twelve hours. Evidence has to be selected, code has to be cleaned, and the explanation has to be precise enough to reproduce.
Writing continuously changes the endgame:
| Approach | Final block | Result |
|---|---|---|
| Write at the end | 10–12 hours reconstructing | Missing evidence, rushed explanations |
| Write as you go | 3–4 hours assembling | Complete findings, time to review |
The report is graded per objective. A finding without proof can score zero regardless of the exploit behind it.
What the exam feels like at hour forty
Quiet, and slightly surreal. The technical work is mostly done; the remaining risk is administrative. This is when candidates discover that a script depends on a session that has expired, or that the proof screenshot does not show the hostname.
The candidates who are calm at hour forty are the ones who tested their exploits from a clean state on day one, and who captured evidence twice: once when it happened, and once in the final review.
Three decisions that define the attempt
- Which target first. Choose the one whose source you can map fastest, not the one that looks most interesting.
- When to stop exploring. Objectives are finite. Exploration is not.
- When to sleep. Before the mistakes start, not after.
What surprised candidates most
Not the difficulty of the vulnerabilities. The amount of writing. OSWE is as much a documentation exam as an exploitation exam, and the candidates who internalise that on day one finish comfortably. The ones who treat the report as an afterthought run out of hours with the technical work already done.
The second surprise is how much source review rewards patience. The intended flaw is usually visible within twenty minutes of focused reading — but only if you are reading the right file, which requires the map you built in the first block.
Where to go next
For the certification overview and study material, see the OSWE certification hub. For exam structure and a preparation plan, read OSWE exam: 48 hours, source review and a custom exploit. For the failure patterns to avoid, see OSWE common mistakes.
Certification names belong to OffSec. This is independent study material, not official course content.
Web application security specialist and bug bounty hunter. OSWA and OSWE certified, focused on web exploitation, API security and source-code review. Writes our web exploitation walkthroughs and reviews report structure for web-focused certifications.