There is no fixed number of years. The honest test is relevance: a role from twelve years ago that used the same tools, or that led directly into your current specialism, earns its place. A role from four years ago that has nothing to do with the job you’re applying for now earns a line, not a paragraph. Age alone is not the reason to cut something, and recency alone is not the reason to keep it.
That said, most CVs that go wrong on this err in the opposite direction from what people expect. The common failure is not too much history. It’s too little trimming of what is kept: every job listed in full detail regardless of how relevant it still is to the role on offer.
Why “10-15 years” became a rule of thumb
The number floating around most guides isn’t arbitrary. It roughly matches how long a skill set or job title tends to stay recognisable before the field around it changes enough that the early years stop reading as evidence of current ability. A decade-old role in a discipline that has since changed tools, methods, or terminology does less work on the page than a shorter, more recent one in the same area.
But treating it as a hard cutoff misreads why the guideline exists. A role from fifteen years ago that is directly relevant to the job posting is worth more than a role from three years ago that isn’t. The number is a rough proxy for relevance, not a substitute for judging it.
What actually happens to the older entries
A CV doesn’t need a binary choice between “include the whole role” and “delete it.” There is a middle option that does most of the real work: compress it. A role that shaped your trajectory but isn’t itself relevant to the target job can sit as a single line, title, employer, and dates, with no bullets under it. It still shows continuity and answers “what were you doing before this,” without spending space defending detail nobody asked for.
Full bullets belong on the roles doing the actual persuading: the ones that match what the posting is asking for. Everything older than that either compresses to a line or drops, depending on whether it still adds context a reader would want.
When dropping something outright is the right call
When it’s unrelated to your current direction and adds nothing to the narrative. A short early-career stint in a different field, before your career direction was set, usually contributes nothing that a reader needs, and holding onto it out of completeness just adds length without adding evidence.
When keeping it would extend the CV past what the role justifies. A CV’s length should match the seniority and depth of the case being made, and an entry from fifteen years back is one of the first places to look when a document is running long for no good reason.
When the gap between it and your current work is wide enough that explaining it would take more space than it’s worth. If a role needs a sentence of context before it makes sense next to your recent work, that’s often a sign it belongs in a line, not a section.
When it’s worth keeping in full, regardless of age
When it’s the origin of your specialism. If a role a decade back is where you started doing the exact kind of work the posting is asking for, its age works in your favour: it shows sustained depth, not a recent pivot.
When the employer or project name still carries weight in your field. Recognition value doesn’t expire on a fixed schedule, and dropping a genuinely relevant credential just to hit a year count removes evidence a reader would have used.
When it’s the only place a required qualification, license, or certification is documented. If a credential only shows up in the context of an older role, cutting that role means cutting the credential too, and that kind of omission reads as a gap even when nothing is actually missing.
What not to conclude
Do not conclude that anything older than ten years should be deleted. Relevance decides that, not age, and a relevant older role beats an irrelevant recent one every time.
Do not conclude that keeping everything in full detail is the safe choice. It isn’t safe, it’s just longer, and length spent on low-relevance history is length not spent on the roles actually doing the persuading.
Do not drop a role that is the only record of a required credential, even if it’s old and otherwise irrelevant. Compress it to a line instead.
The short version
Judge each older role on relevance to the job in front of you, not on how many years have passed. Compress what still adds context but isn’t itself persuasive. Drop what adds neither. Keep in full whatever is still doing real work, no matter how old it is.
We don’t decide what stays and what goes. We build a tool that reshapes the CV itself to match a specific posting, which is what makes a role’s relevance visible in the first place.