How I Check an Animal Hospital Patient Before I Admit It

Last updated 2026-07-14.

Most of the bad admits I have made in Animal Hospital were not because I did not know the anomaly. They were because I made the decision in the wrong order. I looked at the patient, felt fine about it, and opened the door before the photo had finished processing. The anomaly was in the photo. It was always going to be in the photo.

So this post is not a list of tells. The anomaly database already holds all eighteen of the identification signs we track, and the cheat sheet compresses them into something you can keep on a second screen. This is about the sequence, because the sequence is what actually decides whether you catch things.

The order is the whole trick

The routine is: look, photograph, monitor, decide. Four words, and I say them out loud on voice chat during a run because saying them stops me skipping one.

The reason the order matters is that the three categories of anomaly hide in three different places, and they do not overlap. Appearance anomalies are visible at the glass and nowhere else you need. Photo anomalies are invisible at the glass by design. CCTV anomalies are invisible both at the glass and in the photo. A patient that passes your eyes has told you almost nothing about what the camera is going to show.

That means "it looked normal" is not evidence. It is the absence of one kind of evidence. The only sentence that justifies an admit is "it looked normal, the photo matched, and the feed was clean."

Check one: the window

Six of our catalogued tells live here — Three Eyes, Hollow Face, Human Teeth, Creepy Smile, Googly Eyes and Sharp Teeth, and Twitching.

Most of them are loud. A glowing red third eye is not subtle. Empty hollow sockets are not subtle. The one that quietly ruins runs is Human Teeth, because the patient behaves completely normally and the only thing wrong with it is the mouth. It is the reason I now check the mouth on every single patient rather than only the ones that already look strange — by the time a patient looks strange, I did not need the reminder.

Head Banger belongs in a footnote here rather than on that list, and an earlier version of this post got it wrong by counting it as a seventh appearance tell. Every source we re-checked classes it as a hazard event, not a patient you screen and admit. It bangs its head against the reception glass, interacting with it while unprepared drains sanity, and the play is to wait for it out or give it coffee. But you are not deciding whether to let it in, so it is not part of the identification routine.

Check two: the photo

Only three anomalies live in the photo — Different In Photo, Photo Static, and Cursed Photo — and yet this is the step people skip most, because the patient in front of them already looked fine.

Different In Photo is, in my experience, the single hardest tell in the game. The processed image shows altered ears, eyes, colour, or shape versus the animal standing at the glass, and the difference can be small. What fixed it for me was making the comparison mechanical: hold the photo up and check ears, then eyes, then colour, then shape. Four discrete looks. If I try to take in the whole picture at once and ask myself "does this feel different," the answer under shift pressure is always no.

Cursed Photo has a cost attached that the other two do not: looking at the developed photo takes 10 sanity instantly, whether you admit the patient or not. So glance, decide, close. There is no reward for staring at it, and I have watched teammates lose their composure for the rest of a shift doing exactly that.

Check three: the cameras

Eight of the eighteen — Void Body, Deformed Body, Skinwalker, Different Ears, The Starer, Black Eyes, Black Body Red Eyes, and CCTV Twitching — are the ones the cameras reveal. That is nearly half the catalogue sitting behind a step that takes about three seconds.

Void Body is the freebie: the whole body goes solid black on camera and there is nothing to interpret. Skinwalker is the one that punishes you, because a wide mouth full of sharp teeth on the feed is the only warning you get before it walks in and turns hostile inside the hospital.

Different Ears needs a correction I owe you. We used to call it camera-only. It is not: every source we checked places mismatched ears in the photo as well as on the feed. So check the ears twice — once against the photo, once on the camera. If you only ever look for this one on the cameras, the photo version walks straight past you.

The Starer deserves a note because it has a confirmation step built in. It always faces the camera. Switch to another feed and it turns to look at that one too. If you think you have a Starer, that second switch is the confirmation, and it takes a moment.

Black Body Red Eyes is actively hostile while you are still deciding — it drains sanity and zooms the feed in on you. Nothing about that changes the decision. Reject it.

What I do when I am not sure

Two rules, and they have saved more runs than any amount of anomaly memorisation.

The first: one confirmed sign is enough. I do not look for a second one to feel better about the call. If the ears are wrong on camera, the patient is rejected; I do not then go back and re-examine the photo hoping to build a case. Extra confirmation is extra time, and time at the desk is the resource the game is actually taxing.

The second: do not promote a vibe to a verdict. Under pressure it is very tempting to invent a rule mid-shift — "that one moved oddly, I think that's a new anomaly." A team that starts inventing rules starts rejecting clean patients and losing time. If nothing in our list matches, finish the remaining checks rather than acting on a feeling.

That is not the same as claiming our list is complete. It is the set of signs we could corroborate across more than one source, and the game keeps changing. If you see something repeatable that is not in it, trust your eyes and tell us — but do not build a rule out of a single strange-looking patient in the middle of a shift.

That is exactly the gap the anomaly checker is there to close. When I notice a sign and my brain is too loaded to place it, I click the sign and read the verdict. It is not a substitute for looking at the game. It is a way of not having to trust my own memory at the worst possible moment.

How honest this list actually is

Here is the part most guide sites leave out. This list is our reconstruction, not an in-game audit. We did not sit in the game and record every patient that walked up to the glass. We read the public community write-ups, kept a sign when more than one write-up that had clearly not copied the other described it, and dropped anything that showed up in only one place. That is a reasonable way to build a list. It is not the same as knowing, and I would rather say so than let you assume otherwise.

You should treat it accordingly: the checker is a memory aid, not an oracle. A clean result from it does not mean the patient is safe — it means nothing on our list matched what you told it. Your eyes at the glass beat our table every time.

Reading it back on 2026-07-14 against Beebom, Tech Nerdiness and fdaytalk cost us two corrections, both above: Head Banger was filed as an appearance tell when those write-ups call it a hazard, and Different Ears was described as camera-only when the ears can differ in the photo too. Our tier rankings are opinion on top of all that — nothing in the game ranks anomalies, so S-to-C is just our judgement, and the tier list says so right on the page.

The list is also incomplete. Those same write-ups name tells we do not publish — a separate "different eyes" photo variant, a censored-eyes camera tell, a hollow face that only shows on CCTV — because we could not pin them down well enough to put them in front of you as fact.

Animal Hospital is actively updated. If a shift throws something at you that is not in the catalogue, the honest answer is that the catalogue may be behind the game, not that you imagined it. Tell us. That is more useful than us pretending the list is closed.

FAQ

Can I skip the CCTV check if the patient obviously looks fine? No, and "obviously looks fine" is exactly the condition under which the CCTV anomalies were designed to get through. Eight of the eighteen entries are ones the cameras reveal.

Is one anomaly sign really enough to reject? Yes. There is no case in the catalogue where a confirmed sign turns out to be a normal patient, so hunting for a second sign only costs time.

Which anomaly do people miss most? In my runs, Human Teeth and Different In Photo. Both involve a patient that reads as completely normal until you look at one specific thing.

Does rejecting anomalies protect my sanity? Largely, yes. Several of the worst sanity events in the game happen because something hostile got admitted. Catching it at the glass is cheaper than dealing with it in the corridors.

Are there codes that help with any of this? No. Animal Hospital still has no code redemption box at all, which we re-check against the game itself — see the codes page for the current status.