How to Write Photo Captions (and Alt Text)

A good photo caption tells the viewer what the picture cannot: who is in it, what is happening, where and when it was made, and why it matters. Alt text is different. It describes what the picture shows, for someone who cannot see it. Write captions accurately and early (for family photos, while someone still remembers the names), keep alt text short and literal, and store both in the file’s metadata so they travel with the image wherever it goes.

What every caption must answer

A photograph is very good at showing what something looked like and very bad at explaining it. It cannot tell you a person’s name, the date, the town, or what happened a minute before. The caption fills exactly that gap, and the classic checklist still works:

• Who: full names of identifiable people, spelled correctly, with a role or relationship if it matters.
• What: the action or event, in plain words.
• Where: specific enough to find again: venue or street, town, region or country.
• When: a full date where possible, an approximate one otherwise.
• Why: the reason the scene matters or the context a viewer needs to understand it.

Not every caption needs all five in the same depth, but you should consciously decide to leave one out. A useful rule: never describe what the viewer can already see, unless it is ambiguous. “A man in a blue coat” wastes words if the coat is obvious; “Dr. Ana Ruiz (left), the clinic’s director” gives information no one could get from the pixels.

A worked example. Suppose you photograph an elderly man weighing saffron at a market stall. Here is the same caption improving step by step:

1. “Market.” Tells the viewer nothing the picture does not.
2. “A spice seller at the market.” Adds a role but no specifics.
3. “Spice seller Omar Haddad weighs saffron at his stall in the old market, on the last Saturday before the harvest festival.” Now there is a who, what, where, when and a hint of why.
4. Add one line of context if the picture will be seen by strangers: “His family has run the stall for three generations.” Only include it if you checked it.

Notice that version three is one sentence, reads naturally, and contains nothing you could see without being told. That is the target.

Captions for family photos: write the names now

Family archives rarely fail because prints fade. They fail because nobody wrote down who was in them. A box of beautifully preserved photographs of unidentified people is almost worthless to a grandchild, and the only people who can fix that are alive now. Captioning is the single most valuable thing you can do for a family photo archive.

For each photograph, record:

• Full names, left to right, including birth surnames for people who later changed them, and nicknames in quotation marks. “Aunt Bea” means nothing in fifty years; “Beatrice (Bea) Morton, born Clarke, the photographer’s aunt” does.
• A date, even an approximate one. “Summer, about the year Tom started school” is better than nothing, and you can refine it later. Mark guesses clearly with “about” or “probably” so no one mistakes them for fact.
• The place, down to the street or house if known: “Grandpa’s farm on River Road” helps a future reader find it on a map.
• The occasion and the photographer, if known. Who held the camera explains why that person is missing from every picture.
• Your source. “Identified by Mum, from memory” tells future readers how much to trust it.

Sit with the oldest relative you can and go through a stack together. Record the conversation if they agree; the stories that come out are often worth more than the names. When you digitize old photos, type the caption into the scan’s metadata at the same time. For physical prints, write on the back near the edge with a soft pencil, pressing lightly, or with a pen sold specifically as safe for photographs. Avoid ballpoint pens, which can emboss the image from behind, and ordinary felt markers, which can smear onto the next print in the stack. The guide to storing old photos covers the rest.

Documentary and editorial captions: accuracy and fairness

In photojournalism and documentary work, the caption is part of the record, and a wrong caption can do real harm: to the person misidentified, to the story, and to the credibility of every other picture you publish. The standards are strict for good reasons.

• Verify names on the spot. Ask people to write their name in your notebook or type it into your phone. Confirm spelling, age (if relevant) and home town before you leave.
• Describe actions, not inner states. “A woman grieves” is a guess; “A woman stands beside the flooded house, her hand over her mouth” is an observation. Only say what someone felt if they told you, and then attribute it.
• Keep what the picture shows separate from background. Many news organizations write the first sentence in the present tense, describing the moment in the frame, and later sentences in the past tense for background. It makes clear which facts the image itself supports.
• Say when a picture is not what it seems. If a scene was posed, reconstructed, made from several frames or altered beyond routine toning, the caption must say so. The retouching ethics guide explains where the usual lines fall.
• Be fair to the people in the frame. Avoid loaded words, do not imply someone is involved in wrongdoing because they appear near it, and think twice before naming children or people in vulnerable situations.

Crowds need special care. At a festival or protest you cannot identify most people, so describe the event and the visible action instead. And remember that a caption on a crowd picture can attach a label (supporter, protester, rioter) to everyone in it. Choose words that are true of the people actually shown.

Captions for portfolios, exhibitions and books

Outside journalism, captions become quieter. Their job is to support the pictures without explaining them away. A few conventions help:

• Title and caption are different things. A title is an artistic choice (“Low Tide”) and may be poetic or absent. A caption or label is factual. Many photographers use a plain factual title, such as place and date, precisely so it does not steer the viewer.
• Exhibition labels follow a pattern. A typical wall label gives the title, place and date of the photograph, then the print details (process, paper, size, edition). Keep the format identical for every label in a show; inconsistency looks careless.
• Do not tell the viewer what to feel. “A haunting reminder of loneliness” closes the picture down. Let the image do that work and give the facts that enrich it. Longer thoughts belong in an artist statement.
• In books, captions can move. Many photobooks keep the picture pages clean and gather captions in a list at the back, keyed by page number, so the sequence reads without interruption. Others place a short caption under each image. The photo book design guide covers layout; either choice works if it is consistent.
• In a photo essay, captions carry the story between pictures. Each caption should add a fact the next picture builds on, not repeat what the previous one said. The photo essay guide shows how pictures and words interlock.

Captions and alt text are not the same thing

On the web, an image can carry two pieces of text, and people often confuse them.

The caption is visible to everyone, sits beneath or beside the picture, and adds context the picture cannot show. The alt text (alternative text) is normally invisible. It is read aloud by screen readers for people who are blind or have low vision, shown when an image fails to load, and used by search engines to understand what the image contains. It describes what the picture shows.

Take the photograph above. A caption might read: “The fish counter at the Mercat de Sant Josep, known as La Boqueria, in Barcelona.” Its alt text reads: “A market counter heaped with whole fish on ice under warm light, with handwritten price cards and vendors and customers behind.” The caption says where; the alt text says what is there. A screen reader user hears both, so the two should complement each other, not repeat word for word.

How to write good alt text

Alt text is a small piece of writing with a clear purpose: give someone who cannot see the image what a sighted reader gets from it, in the context where it appears.

1. Start with what matters most. Put the main subject first. Screen reader users can stop listening at any point, so the key information should come early.
2. Keep it short. A sentence or two is usually enough. If an image needs a long explanation, such as a chart or a complex scene, put the detail in the caption or the surrounding text.
3. Skip “image of” or “photo of”. Screen readers already announce that it is an image. Only mention the medium when it matters, such as “a hand-tinted portrait” or “an aerial photograph”.
4. Describe, do not interpret. Say what is visible: “A child in a yellow raincoat jumps into a puddle.” Avoid judgments like “a beautiful scene” unless the mood is the point, and then describe what creates it: “soft fog hides the far shore”.
5. Include any text that appears in the image if it matters: signs, headlines, words on a shirt.
6. Match the context. The same photo of a beach needs different alt text on a page about tides (“waves breaking on wet sand at low tide”) than on a page about lens choice (“a wide view of a beach, with the horizon near the top of the frame”).
7. Leave purely decorative images empty. A divider or background texture should have an empty alt attribute so screen readers skip it, rather than a description that wastes the listener’s time.

Weak: “IMG_4032” or “Photo of a market.” Better: “An older man in an apron weighs spices on a brass scale at a crowded market stall.” The better version names the subject, the action and the setting in one sentence.

Store captions in file metadata so they travel

A caption that lives only on a website, in a photo album or in your head is lost the moment the image moves. The fix is to write it into the file itself. Image files can carry descriptive metadata alongside the technical EXIF data your camera records automatically. The widely used standard for descriptive information is the IPTC photo metadata standard, supported by almost every serious photo editor and cataloguing program.

The fields worth filling in:

• Description (also called caption): your full caption. This is the field most programs and news systems read as the caption.
• Headline or title: a short label, a few words long.
• Creator and copyright notice: who made the picture and who owns the rights.
• Date created and location fields: city, region, country and a more specific sublocation.
• Keywords: names, places and subjects for searching later.
• Alt text: newer versions of the standard include a dedicated field for accessibility text, so a description can travel with the file too.

For JPEG, TIFF and DNG files, this information is embedded inside the file. For many camera raw formats, editors store it in a small sidecar file with the same name, which you must keep beside the raw file when you copy or back it up. When you export, check that your export settings include this metadata rather than stripping it.

Work efficiently: select all photos from one event and apply the shared information (event, place, date, your name and copyright) to the whole batch in one step, then add individual names and captions to each frame. A few minutes after every shoot beats a lost weekend years later. Two cautions: many sharing platforms strip some or all metadata on upload, so keep your captioned master files safe; and embedded location and names can reveal more than you intend, which the guide to sharing photos online privately covers.

Common mistakes

• Stating the obvious. “Sunset over the lake” under a picture of a sunset over a lake. Fix: name the lake, the place and the date, or add something the viewer cannot see.
• Guessing names or dates without saying so. A confident wrong caption is worse than none, because it will be repeated. Fix: mark uncertainty with “probably” or “about” and note your source.
• Writing emotions into documentary captions. Fix: describe the visible action and attribute feelings only to what people told you.
• Copying the caption into the alt text. Screen reader users then hear the same words twice and never learn what the picture shows. Fix: caption for context, alt text for what is visible.
• Keeping captions only in a separate document or album. Files get separated from notes. Fix: embed captions in the image metadata, and check they survive export.
• Leaving family photos for “later”. Later often means after the last person who knew the names has gone. Fix: caption ten photos this week.

Try this

Caption ten photos properly (20 to 30 minutes).

1. Choose ten photographs: five recent ones of your own and five older family pictures, printed or scanned.
2. For each, write a one-sentence caption answering who, what, where and when. Mark any guesses.
3. For three of the ten, add one sentence of why: the reason the moment mattered.
4. Write alt text for two of the pictures, in one sentence each, starting with the main subject.
5. Open the files in your photo editor and paste each caption into the description field. Export one copy and check, in a file information window, that the caption is still there.
6. If any older photo stumped you, send it to a relative today and ask who is in it.

Frequently asked questions

How long should a photo caption be?

For most uses, one or two sentences. News captions often run two sentences: the moment, then the context. Exhibition labels are shorter still. Family archive notes can be as long as you like, because detail is the point.

Should a caption be written in the present or past tense?

News and documentary captions commonly describe the action in the frame in the present tense (“A firefighter carries…”), then give background in the past tense. Family and archive captions usually read more naturally in the past tense. Pick one convention and stay consistent.

What is the difference between a caption and alt text?

A caption is visible text that adds context, such as names, places and dates. Alt text is usually hidden and describes what the image shows for people using screen readers or when the image fails to load. Most published photos should have both, and they should not be identical.

Will my captions survive if I upload photos to social media?

Often not. Many platforms remove embedded metadata when you upload. Keep your original captioned files as the master copies, and type the caption into the post as well if it matters.

Do I need permission to name people in my captions?

It depends on the context and on the law where you live and publish, which varies by country. News use generally differs from commercial use, and naming children or people in sensitive situations needs extra care. For anything beyond family and personal use, consult a lawyer in your jurisdiction; the model release guide explains the related question of consent for commercial use.

Related guides

• How to Modify Metadata on Your Pictures: the practical steps for editing the fields described above.
• Family Photo Archive: organizing, preserving and passing on a family collection.
• How to Create a Photo Essay: how captions and sequence carry a story together.
• Photojournalism: the standards that make captions part of the record.
• Photo Book Design: placing captions and text in a printed book.
• Photography Theory: why context and captions change what a picture means.
• How to Organize Your Photos: a filing and keywording system that makes captions searchable.