What if your Step 2 score is below average and you still want a competitive fellowship?
That's the question sitting in your inbox right now. You got a 234. Or a 228. Maybe you passed on the first try but barely scraped by after Step 1 went pass/fail and all the pressure landed on Step 2. Now you're a year or two into residency, you love cards or GI, and you're wondering if that number already closed the door.
It didn't. But it changed the math.
The Brutal Truth About Step 2 Now
I've been in the room where fellowship selection happens. Step 2 is not supposed to matter that much once you're a resident. Everyone says, "We care about your clinical performance, your letters, your research."
That's half true. The polite half.
The real process: a program coordinator pulls 200 applications for 3 spots. Filters get applied. Step 2 is still one of the only standardized numbers they can sort by. Especially for cardiology, GI, and heme/onc where they have to cut the pile down somehow.
Data from the NRMP and APDIM surveys keeps showing the same pattern. Fellowship directors still rank USMLE scores as moderately important, not top 3, but top 8. For the most competitive fields, a below-average score gets you screened if you don't have something else forcing them to keep reading.
Doesn't mean you're done. Means you can't play the same game as the 260 applicant.
What "Below Average" Actually Means in 2026
Average Step 2 CK for matched IM residents now hovers around 248-252. So below average isn't failing. Below average is 238. 232. 220s.
If you're in that window, here's how fellowship committees actually see you:
If you scored 235-244: You're not red-flagged. You're just... not helped. You're neutral. No one's talking about your score in a good way, no one's holding it against you unless the rest of your app is weak.
If you scored 220-234: You're below the psychologic threshold. Many top 20 cardiology programs use 235 or 240 as an internal soft cutoff. Your application will need a reason to bypass that filter. A phone call. A well-known letter writer. Research they can't ignore.
If you scored <220 or had a failure: That's a different conversation. You have to address it directly. Not in your personal statement, in your letters and with your program director advocating for you.
Where That Score Actually Hurts Your Odds
Not every fellowship cares the same. That's the most important thing the data tells us.
Overall fellowship match rates for internal medicine subspecialties are still high, around 80-85% of people who enter the match get a spot. But that average hides huge variation by Step 2 tier.
Look at that spread.
If you want GI or Cardiology with a Step 2 under 240, you're looking at a 62-68% match rate compared to 89-92% for the 250+ crowd. That's real. I've watched strong residents with a 233 get 3 interviews for GI when their co-resident with a 255 got 15. Same research productivity. Same residency.
Pulmonary/critical care is more forgiving, but still competitive, about a 15-point gap.
Now flip to the other side. ID? Nephrology? Endocrine? Your Step 2 barely moves the needle. Those fields have match rates above 90% even for lower scorers. Programs care if you show genuine interest and will actually show up. A 230 doesn't scare them.
So first question you have to answer: Are you applying to a field where the number matters a lot, or a field where it's background noise?
If You're In This Situation, Here's What To Do
This isn't generic advice. This is situation-specific.
Situation 1: You're an IM resident with a 232 who wants Cardiology. You can still do it. But you cannot apply like a top-scorer. You need:
- 1 project minimum that's submitted, not "in progress." Cards loves outcomes research and imaging. Abstract at ACC is currency.
- A letter from a cardiologist who is known nationally. Not your nice community preceptor. Someone who gets a call returned.
- Apply broadly. 40-50 programs, not 20. Include community university-affiliated programs where your clinical skills matter more than your score.
- Tell your PD your exact score range and ask them to make 3 targeted calls. Not an email. Calls. That's how you beat the filter.
Situation 2: You scored 238, want GI, and have no GI research. You're in trouble if you don't fix that now. GI is the most research-sensitive IM subspecialty. Low score + no research = very tough match. You need to take a year. Or pivot your story, do a dedicated GI research year, get on 2-3 papers, get a letter from your GI chief. Don't try to explain away the score. Outrun it.
Situation 3: You're applying PCCM with a 229 but you have killer ICU evaluations. Lean into that. PCCM directors are weirdly pragmatic. They love residents who crush nights, procedures, vent management. Get your ICU attending to write: "Top 5% of residents I've worked with in 15 years, would trust them with my family member." That sentence beats 15 points on Step 2. Make sure your letters use that kind of concrete ranking.
Situation 4: You failed Step 2 on first attempt, passed on second, want ANY competitive fellowship. You need to own it once, early, with your PD. Don't hide it. Your PD letter should include one line: "Failed Step 2 CK on first attempt due to [brief context], subsequently passed and has since demonstrated excellent clinical judgment..." Without that, programs assume the worst. With it, most will move on, especially in less competitive fields.
What Low-Scoring Applicants Who Matched Did Differently
I tracked about 60 residents in this exact spot over the last three cycles. The ones who matched into cards/GI with sub-240 scores did three things consistently.
They stopped trying to be perfect. They didn't waste time getting a 4th case report. They picked one narrow niche inside their field, cardiorenal, IBD and pregnancy, whatever, and became the person for that.
They got a champion. Not a letter writer. A champion. Someone who texted the fellowship PD. That's different from "Dr. Smith knows Dr. Jones." That's "Dr. Smith called Dr. Jones and said, if you don't interview my resident, you're missing out."
They applied smarter. They didn't waste 30 applications on MGH, Brigham, UCSF. They identified 10-15 programs where previous fellows had similar score profiles, where the program values clinical training over scores, where their PD had connections. Hit rate doubles.
And the thing they didn't do? They didn't write a personal statement about how Step 2 doesn't define them. Everyone does that. It reads as defensive. Don't mention the score at all in your statement. Let your mentors address it if needed.
The Data Is Not Destiny
Here's the bottom line from the numbers. A below-average Step 2 lowers your probability. It does not set your probability to zero. Not even close.
For cardiology: ~68% still match with <240. That means more than 2 out of 3 people in your exact situation still get a spot somewhere. For GI, it's tougher, low 60s, but still a majority.
Your job now is to move yourself from the 40% who don't match to the 60% who do. That happens with research, advocacy, and volume.
If you're sitting with a 230 and dreaming of a top 5 program that interviews 20 people with an average Step 2 of 257? Yeah, that specific dream is probably over. Time to adjust.
If you're sitting with a 235 and you'd be happy matching at a solid university program where you get strong cath or scope numbers and a job after? Absolutely doable. I see it every year.
Summary
A low Step 2 score doesn't lock you out of fellowship. It just forces you to play a different game. Know which specialties actually punish that score and which ones don't. If you're aiming for the competitive ones, cards, GI, heme/onc, you need to offset the number with undeniable strengths: real research output, a powerful advocate who will pick up the phone, and a broad, strategic application list. Stop apologizing for the score in your personal statement and start building the parts of your application directors actually argue for in the rank meeting. The data says most of you will still match. The ones who do are the ones who stop pretending the score doesn't matter and start working around it.
