Resume screening compares each application with the job's written requirements to decide who advances. Here is a two-pass process, a weighted scorecard, what to look for, and the legal guardrails for human and AI screening.
- Write your must-have and preferred criteria before you open the first resume, and keep them tied to the actual job.
- Use two passes: a quick yes/no check on must-haves, then a scored review of the resumes that remain.
- A simple weighted scorecard makes screening faster, fairer and easier to defend than gut feel.
- Any screening rule, whether applied by a person, an ATS filter or an AI tool, must be job-related if it screens out a protected group at a higher rate.
- Keep application records for at least one year under federal EEOC rules, and longer if a charge is filed.
In this article
Resume screening is the first review of job applications, where you compare each resume against the role’s written requirements and decide who moves on to a phone screen or interview. Done well, it is fast, consistent and documented: you set criteria before reading, score every resume the same way, and keep a record of why each person advanced or didn’t.
What resume screening is, and how it differs from CV screening
Resume screening sits between the job posting and the first conversation with a candidate. Its job is narrow: remove applications that clearly don’t meet the role’s core requirements and rank the rest, so your interview time goes to the most likely fits. It is one part of broader candidate screening, which also includes phone screens, assessments and background checks.
CV screening is the same task with a different document. U.S. resumes are usually one to two pages and tailored to the role. A CV, common in academia, medicine and outside the U.S., runs longer and lists everything. The method below works for both.
Screening can be manual, automated through an applicant tracking system, or a mix. Most teams use knockout questions or filters for hard requirements, then a person reviews what’s left.
Set screening criteria before you read any resumes
Criteria that shift mid-review are the main cause of inconsistent, hard-to-defend screening. Meet with the hiring manager for 15 minutes and agree on two short lists.
- Must-haves (knockouts): requirements the person truly cannot do the job without. Examples: an active RN license in the state, a CDL Class A, legal authorization to work, availability for the posted shift. Keep this list to three to five items.
- Preferred (scored): things that make a candidate stronger but that a capable person could learn or work around. Examples: experience with a specific software, industry background, bilingual skills.
For each criterion, ask one question: if a candidate lacked this, could they still do the job well within a reasonable ramp-up period? If yes, it belongs on the preferred list, not the knockout list. Degree requirements and fixed years-of-experience minimums often fail this test.
Tip: Copy the criteria straight from the job description. If something matters enough to screen on, it should be in the posting; if it’s in the posting but nobody would screen on it, cut it from the posting. Our guide on how to write job descriptions covers this.
The resume screening process, step by step
- Apply knockout criteria first. Check each resume against the must-haves only. This is a yes/no decision that should take well under a minute per resume. Use ATS knockout questions for objective items such as licenses and work authorization.
- Score the remaining resumes. Rate each one on the same weighted criteria (template below). Read for evidence, not keywords: what did the person do, at what scale, with what result?
- Sort into three groups. Advance (strong scores), hold (borderline, worth a short phone screen if the top group is thin) and decline.
- Calibrate on the first ten. Have the hiring manager score the same ten resumes you did. If your scores differ a lot, talk through why and adjust the criteria before you continue.
- Record the reason for each decision. One line in the ATS is enough, for example “no active state license” or “score 2.1, below advance threshold of 3.5.”
- Respond to candidates. Move advancing candidates to a phone screen within a few days and send declined applicants a short, respectful note. Our candidate rejection email templates can help.
How to evaluate a resume: what to look for
Once a resume clears the knockouts, you are judging how well the person’s past work predicts success in this role. These are the signals worth your time, and the ones to set aside.
| Look for | What it tells you |
|---|---|
| Relevant responsibilities at a similar scale | Whether they’ve done this kind of work, not just held a similar title |
| Specific results (numbers, outcomes, scope) | Impact, such as “cut month-end close from 8 days to 5” rather than “responsible for close” |
| Required tools, licenses and certifications | Readiness to start without extra training |
| Growth in responsibility over time | Promotions, larger teams or budgets, harder projects |
| Transferable skills from other industries | Capable candidates you’d miss with a narrow title match |
| Clear, organized writing | Relevant mainly when the role is writing- or detail-heavy |
Set these aside or treat them with care:
- Employment gaps. Gaps often reflect caregiving, health, layoffs or education. If a gap matters, ask about it in the phone screen rather than rejecting on the resume.
- Short tenures. Contract, temporary and project work produces short stints by design. Check whether the roles were contract before reading them as job-hopping.
- School names and graduation years. Prestige rarely predicts job performance, and graduation dates are a proxy for age.
- Names, photos, addresses and personal details. They tell you nothing about job fit and invite bias. Some teams hide them during the first pass.
- Formatting and design. Unless the job is in design, a plain resume isn’t a weak candidate.
A resume screening scorecard you can use today
Score each preferred criterion from 1 (no evidence) to 5 (strong evidence), multiply by its weight, and add up the results. Weights should total 100%.
| Criterion (example: accounts payable specialist) | Weight | Score (1-5) | Weighted |
|---|---|---|---|
| Full-cycle AP experience (invoice to payment) | 35% | 4 | 1.40 |
| ERP experience (NetSuite or similar) | 25% | 3 | 0.75 |
| Volume handled (invoices per month, vendors) | 20% | 5 | 1.00 |
| Process improvement or reconciliation results | 20% | 2 | 0.40 |
| Total | 100% | 3.55 |
Pick a threshold before you start, for example advance at 3.5 or above and hold at 3.0 to 3.4. In this example, the candidate scores 3.55 and advances. Use anchors so “3” means the same thing to every reviewer: 1 = not mentioned, 3 = done it in a supporting role, 5 = owned it with clear results.
The same structure carries into interviews. Our interview scorecard templates and guide to shortlisting candidates pick up where screening ends.
Screening resumes with an ATS or AI tools
Software helps most with volume and consistency. Knockout questions, structured application fields and saved filters handle objective criteria well. Keyword matching and AI ranking are weaker: they can reward candidates who copy the job ad and miss people who describe the same skills in different words.
- Use automation for objective must-haves (license, shift availability, work authorization), not judgment calls.
- Review a sample of auto-rejected resumes each month to see who the filters are cutting.
- Ask vendors what data their ranking uses, whether it has been tested for adverse impact, and whether you can see why a candidate was scored low.
- Track pass-through rates by stage so you can spot a filter that screens out one group much more than others.
- Keep a human decision-maker before any rejection that isn’t based on a clear knockout.
For a broader view of what these tools do well and badly, see AI in recruitment.
Legal guardrails for resume screening
Screening is a selection decision, so U.S. anti-discrimination law applies to it, whether a recruiter, a hiring manager or software makes the cut. The main rules to know, as of October 2026:
- Disparate impact. Under Title VII, a neutral-looking practice that screens out people based on race, color, religion, sex or national origin at a higher rate is unlawful unless the employer shows it is “job related for the position in question and consistent with business necessity” (42 U.S.C. 2000e-2(k)). This applies to degree requirements, keyword filters and AI scoring alike.
- Protected traits. The EEOC lists race, color, religion, sex (including pregnancy, sexual orientation and transgender status), national origin, age (40 or older), disability and genetic information. The EEOC also says employers should not ask applicants for a photograph and are prohibited from making disability-related inquiries before a job offer.
- Age signals. The EEOC gives “recent college graduates” as an example of ad language that may discourage people over 40 from applying. Screening out candidates for “too much experience” or old graduation dates carries the same risk.
- Recordkeeping. Federal rules require employers to keep application forms and other hiring records for one year from the date the record was made or the action taken, whichever is later, and to keep relevant records until final disposition if a discrimination charge is filed (29 CFR 1602.14). Your screening notes and scores are part of that record.
- New York City. Local Law 144 bars employers and employment agencies from using an automated employment decision tool unless it has had a bias audit within the past year, a summary of the results is posted, and candidates get notice 10 business days before use. Enforcement began July 5, 2023 (NYC DCWP).
- Illinois. An amendment to the Illinois Human Rights Act (HB 3773), effective January 1, 2026, prohibits using AI in hiring decisions in a way that has a discriminatory effect, bans using ZIP codes as a proxy for protected classes, and requires notice when AI is used (Taft summary).
Other states and cities have their own rules on criminal history questions, salary history and automated tools, and they change often. Check the rules where your candidates live and work. Our guides to illegal interview questions and reducing hiring bias cover related steps.
This is general information, not legal advice. Check with employment counsel for your situation.
Common resume screening mistakes
- Too many knockouts. Each extra must-have shrinks the pool. If you’re getting few qualified applicants, review the knockout list first.
- Screening on title instead of work. “Office manager” at one company can mean bookkeeping, HR and facilities; at another, reception only.
- Reviewing tired. Scores drift late in a long session. Screen in batches of 20 to 30 and recheck a few early resumes at the end.
- No feedback loop. Compare screening scores with interview results and 90-day performance, then adjust weights.
- Slow response. Strong candidates are often in several processes at once. Screen new applicants at least twice a week.
How Careerscape can help
Careerscape is a U.S. staffing and recruiting firm. If application volume is more than your team can screen well, our recruiters can source, screen and present a shortlist through direct hire search, or take on the full front end of hiring through recruitment process outsourcing.
Frequently asked questions
How long should it take to screen a resume?
The knockout pass should take well under a minute per resume, because you’re only checking a few must-have requirements. A scored review of a resume that clears the knockouts usually takes a few minutes. If screening takes much longer, your criteria are probably too vague or too numerous. Tighten them with the hiring manager before reviewing more.
What do recruiters look for when screening resumes?
Recruiters first check hard requirements such as licenses, location, schedule and work authorization. Then they look for evidence of relevant work at a similar scale, measurable results, required tools and growth in responsibility. Good recruiters read for what the person actually did rather than matching job titles or keywords, and they avoid judging on school names, photos or personal details.
Is it legal to use AI to screen resumes?
Yes, but the employer remains responsible for the outcome. Under Title VII, a tool that screens out a protected group at a higher rate must be job-related and consistent with business necessity. Some places add specific rules: New York City requires annual bias audits and candidate notice for automated tools, and Illinois requires notice and bans discriminatory AI use from January 1, 2026.
How long do employers have to keep resumes?
Under federal EEOC rules (29 CFR 1602.14), employers must keep application forms and hiring records for one year from when the record was made or the personnel action happened, whichever is later. If a discrimination charge is filed, relevant records must be kept until the matter is fully resolved. Some state laws and federal contractor rules require longer retention.
What is the difference between resume screening and candidate screening?
Resume screening is the review of the written application only. Candidate screening is the broader set of early checks, which can include resume screening, knockout questions, phone screens, skills assessments, reference checks and background checks. Resume screening is usually the first and fastest step in that sequence.
- Cornell LII: 42 U.S. Code 2000e-2, Unlawful employment practices
- U.S. EEOC: Prohibited Employment Policies/Practices
- eCFR: 29 CFR 1602.14, Preservation of records made or kept
- NYC Department of Consumer and Worker Protection: Automated Employment Decision Tools
- Taft: New Illinois Law Restricts Use of AI in Employment Practices