- The three sizes that cover you
- File size and format, the boring but necessary bit
- The mobile crop most people never check
- What this means for carousels and document posts
- A quick step-by-step for sizing a photo before you post
- Photos aren't the only image decision that matters
- Where post photos fit against post images generally
- One more thing worth knowing
- Frequently asked questions
- Useful references
The short version: upload your LinkedIn post photos at 1080 x 1080 pixels for square, 1080 x 1350 for vertical, or 1200 x 627 for landscape, keep the file under 5MB, and stop obsessing over perfect crops because your phone camera probably outperforms your Canva template anyway. LinkedIn recompresses and recrops almost everything on mobile, so the size matters less than most people think, and what shows up on screen matters far more.
The three sizes that cover you
I get asked this constantly by clients who've just spent forty minutes in a design tool trying to get one photo "right" before a launch post. Here's what I tell them, and what I use myself.
- Square: 1080 x 1080 pixels (1:1 ratio). This is my default for almost everything. It displays fully on desktop and mobile without LinkedIn slicing off the top or bottom of someone's face.
- Vertical: 1080 x 1350 pixels (4:5 ratio). This is the one to use when you want to dominate the mobile feed. It takes up noticeably more vertical scroll space than a square post, which means someone has to work harder to scroll past you.
- Landscape: 1200 x 627 pixels (1.91:1 ratio). This one is mostly for link-share posts where LinkedIn pulls a preview image from a URL. If you're uploading a photo directly rather than sharing a link, I'd avoid pure landscape unless it's a screenshot or a wide venue shot where the width is the point.
LinkedIn will technically accept images as large as 7680 x 4320 pixels, but there's no benefit to uploading at that size. It compresses everything down anyway, and a bigger file just means a slower upload and a bigger dent in your data allowance if you're posting from your phone at an airport, which, let's be honest, is where half of us are writing our posts from.
File size and format, the boring but necessary bit
Keep your image file under 5MB. JPG and PNG both work fine, GIFs are supported too if you're doing something animated, but static JPG is what I use for 90% of my own posts because it compresses cleanly without the file bloat PNG sometimes carries. If you're exporting from Photoshop or Canva, export at "web" quality rather than print quality. Print-quality exports are often 20 to 40MB, which LinkedIn will reject or crush so hard the image looks muddy by the time it's live.
The mobile crop most people never check
Here's the bit that costs people reach, and it's not about pixel dimensions at all. Roughly four out of five LinkedIn sessions happen on a phone. When someone views your post on mobile, LinkedIn doesn't show your image at the ratio you uploaded it, it centre-crops to fit the app's feed window. Upload a wide landscape photo of you speaking on a panel and there's a real chance the app slices off your head on one side, or crops the slide behind you into meaningless letters.
I found this out the hard way at a conference in Tel Aviv two years ago. I'd taken a proper landscape photo, a nice wide shot of the stage, me mid-sentence, screen behind me with our session title on it. Looked great on desktop. On mobile, the crop cut off the entire slide and left me looking like I was talking to nobody in an empty room. The post still did fine, but not as well as the raw vertical photo my colleague took on her phone from the audience five minutes later, uploaded with zero editing. That one got roughly 14,200 impressions and an 8.1% engagement rate. My carefully composed landscape shot from the same event got about 3,800 impressions and 1.9% engagement. Same event, same day, same audience. The difference wasn't quality, it was orientation and, I'll admit, authenticity. Hers looked like a person at a conference. Mine looked like a press photo.
That's the part most posts about image sizing skip over, because it's an uncomfortable one for anyone selling design services: getting the pixels technically correct does not make a post perform. A slightly wonky, real, phone-shot photo will usually beat a polished, on-brand, perfectly-sized graphic, because LinkedIn's feed rewards content that doesn't look like it was made to sell you something. Nail the dimensions by all means, it takes thirty seconds and there's no excuse not to, but don't mistake correct sizing for a strategy.
What this means for carousels and document posts
If you're building a carousel (LinkedIn calls these "document posts"), the sizing rules are slightly different because you're uploading a PDF rather than a single image. I build mine at 1080 x 1080 pixels per slide for square carousels, or 1080 x 1350 if I want each slide to feel bigger on mobile. Keep every slide the same dimension, because LinkedIn will stretch or squash inconsistent slides to match the first one in the file, and that's an ugly way to lose someone on slide two.
For text-heavy carousel slides, keep your safe zone (the area guaranteed not to get cropped or covered by UI elements like the like/comment bar) roughly 100 pixels in from every edge. I learned this after a client's call-to-action text on the final slide kept getting half-covered by LinkedIn's own comment box preview. Small margin, easy fix, mildly embarrassing to discover after the post already had 40 comments.
A quick step-by-step for sizing a photo before you post
- Decide what the photo is: a real moment (event, office, screenshot) or a designed graphic. This decides your ratio more than any rule does.
- For real photos, crop to square (1080 x 1080) or vertical (1080 x 1350). Avoid wide landscape unless the width itself is the story.
- For designed graphics, use 1080 x 1080 and keep any text at least 100 pixels from every edge.
- Export as JPG, check the file is under 5MB.
- Before you post, preview it on your own phone, not just desktop. This single step catches 90% of crop disasters.
- Post, then check how it renders in the mobile feed within the first five minutes. If a face or headline is cut off, delete and re-upload rather than leaving it live.
That last step feels fussy, but it takes less time than writing the caption did.
Photos aren't the only image decision that matters
Post photos get most of the attention because people post more often than they update their profile, but your profile picture is doing quieter, longer-term work every single time someone clicks through from one of your posts. I've written before about whether your LinkedIn profile picture matters, and the short answer is that it matters more than most people budget time for, because it's the first thing someone checks when your post photo has already earned their attention. If you haven't checked your own profile picture recently, the exact pixel size for a LinkedIn profile picture is worth five minutes of your time, and it's not the same number as your post images.
There's a difference too between what your profile picture technically is (a headshot, sized correctly) and what it signals about your business before anyone reads a word you've written. A blurry, out-of-date, or oddly-cropped photo tells people something about you whether you meant it to or not.
Where post photos fit against post images generally
I've covered the broader question of ideal sizing for LinkedIn post images in general elsewhere, covering graphics, infographics, and quote cards specifically. This piece is narrower on purpose: real photos, of real people, at real events, behave differently to designed images, and they deserve their own rules rather than being lumped in with a branded template. If you're choosing between the two for a specific post, ask which one a stranger scrolling at 7am on their phone would stop for. In my experience, that's the photo nine times out of ten.
One more thing worth knowing
If you're managing image sizing across more than one platform, it never stops being a small headache, because every platform has its own rules and most of them change without much warning. I ran into exactly this with WhatsApp Business recently and ended up writing about it because there isn't a fixed answer for that platform the way there is for LinkedIn, which you can read in what size your WhatsApp Business banner image should be. LinkedIn, at least, gives you a firm number to work with. Use it, but don't treat it as the whole job.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
Related reading: how to unlike a post on linkedin.
Related reading: does linkedin automatically post work anniversary.
Related reading: editing a linkedin post after publishing.
Frequently asked questions
What is the best image size for a single LinkedIn post photo?
1080 x 1080 pixels (square) works for almost every situation and avoids mobile cropping issues. Use 1080 x 1350 (vertical) if you want the photo to take up more space in the mobile feed, which increases the chance someone pauses to look before scrolling past.
Does LinkedIn crop or resize photos after you upload them?
Yes. LinkedIn compresses every uploaded image and, on mobile, crops it to fit the feed window regardless of the original ratio you uploaded. Wide landscape photos are the most likely to lose important detail, like a face or a headline, at the edges.
What is the maximum file size for a LinkedIn post photo?
Keep files under 5MB. JPG and PNG are both supported, but JPG usually compresses more cleanly for photos without visible quality loss. Export at web quality rather than print quality to avoid oversized files.
Should I use a real photo or a designed graphic for a LinkedIn post?
Real, unpolished photos of people, events, or screenshots typically outperform designed graphics on LinkedIn because they don't read as promotional content in the feed. Get your sizing right on either, but don't assume a more polished graphic will beat a genuine phone photo, because in my own testing it usually doesn't.
