Quick answer
Portrait screenshots show three across in App Store search results, while a single landscape screenshot takes that whole row on its own. That means landscape buys you one large canvas instead of three small ones, which is a good trade for landscape games and a poor one for almost every portrait app. Pick the orientation your app actually runs in, because a portrait app shown in landscape frames looks wrong before anyone reads the caption.
Short Answer
> Quick answer: Use the orientation your app runs in. That sounds obvious and it's still the most common mistake, because designers pick the orientation that suits their layout rather than the one that matches the product.
The mechanic worth knowing is the search results row. Portrait screenshots appear three across in App Store search, so a browsing user sees three of your images before tapping anything. A single landscape screenshot fills that same row on its own. So landscape gives you one big statement and portrait gives you a three panel story, and which is better depends entirely on whether your pitch needs width or sequence.
The mechanic worth knowing is the search results row. Portrait screenshots appear three across in App Store search, so a browsing user sees three of your images before tapping anything. A single landscape screenshot fills that same row on its own. So landscape gives you one big statement and portrait gives you a three panel story, and which is better depends entirely on whether your pitch needs width or sequence.
The Search Results Row Is the Real Difference
Everything else about this decision is downstream of what happens before the tap.
In App Store search results, a portrait listing shows the first three screenshots side by side at small size. A landscape listing shows the first one, alone, at roughly triple the width. Same vertical space, completely different amount of information delivered.
Three portrait panels let you tell a small story. Problem, action, result. Or three distinct features. The user reads left to right and gets a sense of the app's shape without ever opening your page, which is genuinely valuable when most people never do open it.
One landscape panel gives you a single frame with real width, which suits a game world, a dashboard, a timeline or anything where the interest is horizontal. Cramming that into a portrait phone frame at thumbnail size wastes it.
The failure case is a portrait app that ships landscape screenshots to grab the wider slot. You get one image showing a narrow phone floating in a wide empty background, and the visual message is that there wasn't enough app to fill the frame. Our first screenshot guide covers how much that first impression carries.
In App Store search results, a portrait listing shows the first three screenshots side by side at small size. A landscape listing shows the first one, alone, at roughly triple the width. Same vertical space, completely different amount of information delivered.
Three portrait panels let you tell a small story. Problem, action, result. Or three distinct features. The user reads left to right and gets a sense of the app's shape without ever opening your page, which is genuinely valuable when most people never do open it.
One landscape panel gives you a single frame with real width, which suits a game world, a dashboard, a timeline or anything where the interest is horizontal. Cramming that into a portrait phone frame at thumbnail size wastes it.
The failure case is a portrait app that ships landscape screenshots to grab the wider slot. You get one image showing a narrow phone floating in a wide empty background, and the visual message is that there wasn't enough app to fill the frame. Our first screenshot guide covers how much that first impression carries.
Matching the Orientation to the App
The rule is simple and the edge cases are where people get into trouble.
If your app locks to portrait, use portrait screenshots. Everything else is a mismatch between what the store shows and what the user gets after installing, and that gap produces uninstalls even when it produces installs.
If your app locks to landscape, which mostly means games and some video and music tools, use landscape. Trying to force a landscape game into portrait frames means either letterboxing the gameplay into a strip or cropping out the parts of the screen that made it look good.
If your app supports both, pick the one people actually use. Check your own analytics rather than guessing, because the answer is frequently lopsided even for apps that technically support rotation. Then build the set in that orientation and let the store handle the rest.
Tablet listings are the one place to think again. The same app frequently makes more sense in landscape on a tablet even when it's portrait on a phone, and both stores let you upload a separate set per device class. Our tablet and foldable guide goes through the sizes Play expects.
If your app locks to portrait, use portrait screenshots. Everything else is a mismatch between what the store shows and what the user gets after installing, and that gap produces uninstalls even when it produces installs.
If your app locks to landscape, which mostly means games and some video and music tools, use landscape. Trying to force a landscape game into portrait frames means either letterboxing the gameplay into a strip or cropping out the parts of the screen that made it look good.
If your app supports both, pick the one people actually use. Check your own analytics rather than guessing, because the answer is frequently lopsided even for apps that technically support rotation. Then build the set in that orientation and let the store handle the rest.
Tablet listings are the one place to think again. The same app frequently makes more sense in landscape on a tablet even when it's portrait on a phone, and both stores let you upload a separate set per device class. Our tablet and foldable guide goes through the sizes Play expects.
What Each Orientation Costs You in Design
The design constraints are genuinely different, and knowing them upfront saves a redesign.
Portrait gives you vertical room for a caption above the device and a tall device frame below it. That's the layout almost every good listing uses, because the caption reads first and the screenshot supports it. It's a proven arrangement and it scales down to thumbnail size well, since the caption stays legible.
Landscape has almost no vertical room. Once a wide device frame is placed, there's very little space left above or below for text, so captions end up overlaid on the image or squeezed into a corner. Overlaid text on a busy game scene is hard to keep readable, and readability at small sizes is the whole game here.
The workaround for landscape is to design the caption into the composition rather than bolting it on. Reserve a darker region of the scene for text, or use a solid band that's part of the design rather than a label sitting on top of it.
Portrait's weakness is the opposite. A tall thin frame at thumbnail size shows very little detail, so any screenshot relying on fine UI elements reads as texture rather than content. That's why the caption carries the meaning in portrait listings and the image mostly carries the mood.
There's a middle option people forget, which is zooming the screenshot inside the portrait frame. Instead of showing the whole app screen shrunk to nothing, crop to the one component that matters and let it fill the frame. The device outline still signals that it's an app, and the content is finally large enough to read. This is the single easiest upgrade to a portrait set that looks busy at small size.
The equivalent trick in landscape is the opposite. Pull back rather than in, and let the width do the work, because a landscape frame cropped tight loses the only advantage it had over portrait in the first place.
Portrait gives you vertical room for a caption above the device and a tall device frame below it. That's the layout almost every good listing uses, because the caption reads first and the screenshot supports it. It's a proven arrangement and it scales down to thumbnail size well, since the caption stays legible.
Landscape has almost no vertical room. Once a wide device frame is placed, there's very little space left above or below for text, so captions end up overlaid on the image or squeezed into a corner. Overlaid text on a busy game scene is hard to keep readable, and readability at small sizes is the whole game here.
The workaround for landscape is to design the caption into the composition rather than bolting it on. Reserve a darker region of the scene for text, or use a solid band that's part of the design rather than a label sitting on top of it.
Portrait's weakness is the opposite. A tall thin frame at thumbnail size shows very little detail, so any screenshot relying on fine UI elements reads as texture rather than content. That's why the caption carries the meaning in portrait listings and the image mostly carries the mood.
There's a middle option people forget, which is zooming the screenshot inside the portrait frame. Instead of showing the whole app screen shrunk to nothing, crop to the one component that matters and let it fill the frame. The device outline still signals that it's an app, and the content is finally large enough to read. This is the single easiest upgrade to a portrait set that looks busy at small size.
The equivalent trick in landscape is the opposite. Pull back rather than in, and let the width do the work, because a landscape frame cropped tight loses the only advantage it had over portrait in the first place.
Mixing Orientations, and Why Not To
Both stores technically allow a mixed set, and it's almost always a mistake.
Mixed sets break the carousel rhythm. Swiping from a tall image to a wide one changes the layout mid scroll, and the eye has to reset each time rather than flowing through a sequence. A listing that reads as one designed set converts better than one that reads as an archive.
It also breaks the search preview logic. The row treats your first images according to their orientation, so mixing changes what gets shown in ways that are hard to predict and hard to test. If your first screenshot is landscape and your second and third are portrait, what a browsing user sees is not something you can reason about confidently, and an A/B test on a listing that behaves unpredictably in search wastes the traffic you spent on it.
The narrow exception is an app with a genuinely dual mode, like a video editor with a portrait capture flow and a landscape editing timeline. Even then, keeping the first three consistent and putting the odd one later is safer than alternating.
If you're tempted to mix because one feature only looks good sideways, the better fix is usually to crop or reframe that feature so it works in your primary orientation. A tightly cropped detail in portrait beats a full landscape shot that disrupts the set.
I've found the tidiest test is to lay the whole set out side by side at the size the store actually renders them, not at full resolution. Problems that are invisible at 100 percent are obvious at thumbnail scale, and thumbnail scale is where the decision gets made.
Mixed sets break the carousel rhythm. Swiping from a tall image to a wide one changes the layout mid scroll, and the eye has to reset each time rather than flowing through a sequence. A listing that reads as one designed set converts better than one that reads as an archive.
It also breaks the search preview logic. The row treats your first images according to their orientation, so mixing changes what gets shown in ways that are hard to predict and hard to test. If your first screenshot is landscape and your second and third are portrait, what a browsing user sees is not something you can reason about confidently, and an A/B test on a listing that behaves unpredictably in search wastes the traffic you spent on it.
The narrow exception is an app with a genuinely dual mode, like a video editor with a portrait capture flow and a landscape editing timeline. Even then, keeping the first three consistent and putting the odd one later is safer than alternating.
If you're tempted to mix because one feature only looks good sideways, the better fix is usually to crop or reframe that feature so it works in your primary orientation. A tightly cropped detail in portrait beats a full landscape shot that disrupts the set.
I've found the tidiest test is to lay the whole set out side by side at the size the store actually renders them, not at full resolution. Problems that are invisible at 100 percent are obvious at thumbnail scale, and thumbnail scale is where the decision gets made.
Building the Set Either Way
Whichever orientation you land on, a few things make the set hold together.
Use one device frame throughout. Mixing a bezel less frame with an older one, or switching between a phone and a tablet mid set, reads as carelessness even to people who couldn't name what changed. Pick one from the frames library and stay with it across the whole listing.
Keep the caption in the same position on every image. Moving it between top and bottom makes the sequence feel unrelated, and a fixed position lets the eye read the captions as a single sentence while swiping.
Hold your type size constant and check it at store scale. The most common readability failure is a caption sized for a design canvas rather than for a search result thumbnail, and it's invisible until you shrink the image.
Export at the exact dimensions each store wants rather than letting an upload resize for you. Apple's screenshot specifications list the required pixel sizes per device class, and Play publishes its own.
Then build the whole thing in one pass in the editor rather than exporting images over several sessions. A set made in one sitting looks like a set. One assembled across three weeks looks like three weeks of decisions, and that shows in the one place you can least afford it.
Use one device frame throughout. Mixing a bezel less frame with an older one, or switching between a phone and a tablet mid set, reads as carelessness even to people who couldn't name what changed. Pick one from the frames library and stay with it across the whole listing.
Keep the caption in the same position on every image. Moving it between top and bottom makes the sequence feel unrelated, and a fixed position lets the eye read the captions as a single sentence while swiping.
Hold your type size constant and check it at store scale. The most common readability failure is a caption sized for a design canvas rather than for a search result thumbnail, and it's invisible until you shrink the image.
Export at the exact dimensions each store wants rather than letting an upload resize for you. Apple's screenshot specifications list the required pixel sizes per device class, and Play publishes its own.
Then build the whole thing in one pass in the editor rather than exporting images over several sessions. A set made in one sitting looks like a set. One assembled across three weeks looks like three weeks of decisions, and that shows in the one place you can least afford it.
Frequently asked questions
Should I use portrait or landscape app store screenshots?
Use whichever orientation your app actually runs in. Portrait for portrait apps, landscape for landscape games and video tools. A mismatch is visible immediately and produces installs that uninstall.
How many screenshots show in App Store search results?
Three portrait screenshots appear side by side, or one landscape screenshot filling the same row on its own. That is why landscape gives you a single large canvas and portrait gives you a three panel sequence.
Can I mix portrait and landscape screenshots?
Both stores allow it and it rarely helps. Mixing breaks the carousel rhythm and makes the search preview harder to predict. If one feature only works sideways, crop it to fit your primary orientation instead.
Do landscape screenshots convert better?
Only for apps that run in landscape. The wider slot is an advantage when your content is horizontal, and a liability when it shows a narrow phone floating in empty space. Our [first screenshot guide](/blog/first-screenshot-app-store-conversion-2026) covers what that slot needs to do.
Should tablet screenshots use a different orientation?
Often yes. Many apps are used in portrait on a phone and landscape on a tablet, and both stores accept a separate set per device class, so matching each one to real usage is worth the extra export.
Build the mockup in your browser.
Drop a screenshot into a real device frame and export at the exact store size — free, no signup.