Midjourney Prompts for Fashion Photography: A Director's Guide

Midjourney is the AI image model most used by fashion photographers, creative directors, and brand teams for concept exploration and campaign visualisation. Its training data skews heavily toward editorial and art photography, which makes it particularly responsive to the visual language of fashion, when you know how to speak it.

The challenge is that most fashion teams approach Midjourney with the intuitions of a copywriter rather than a creative director. They write descriptive prompts and get descriptive images, technically competent, visually generic. The difference between that and editorial quality comes down to prompt structure.

This guide covers the specific prompt structures, parameter choices, and vocabulary decisions that produce consistent, editorial-quality fashion photography in Midjourney.

How Midjourney Reads Fashion Prompts

Midjourney processes prompts as a weighted sequence of concepts. The terms you use early in the prompt carry more weight than terms used later. This matters enormously for fashion: if you lead with the garment description, you get a garment-forward image. If you lead with the atmosphere, you get an atmosphere-forward image.

For editorial fashion photography specifically, atmosphere and light should almost always lead. The garment exists within a visual world, it doesn't define the visual world. This is the fundamental mistake most fashion teams make when they first start using Midjourney.

The Prompt Architecture for Fashion

Layer 1: Atmosphere and emotional register

Open with the feeling and tonal register you're after. This is not a description of what is in the image, it's a description of what the image feels like. Words like "melancholic," "austere," "restrained," "charged," "desolate" communicate the emotional world to Midjourney in a way that shapes everything that follows.

Layer 2: Environment and spatial logic

Specify the space and how the subject relates to it. "Brutalist concrete interior, late afternoon" gives Midjourney very specific information about architectural texture, scale, and light quality. "Minimalist white studio" gives it much less, and tends to produce generic images.

Layer 3: Subject and casting

Describe the subject in terms of presence and energy rather than physical description. "A woman in her late thirties, stillness in her body, no performed emotion" is more useful than "beautiful woman, fashion model." Midjourney responds to character descriptors, not demographic ones.

Layer 4: Light

Lighting is the single most important variable in editorial fashion photography, and it's the most commonly underspecified element in AI prompts. Go beyond "soft light" or "dramatic lighting." Specify the quality, direction, and colour of the light. "Oblique window light, cool, slight underexposure" gives Midjourney something specific to work with.

Layer 5: Technical and reference vocabulary

Specify the photographic format, lens, and any reference photographers or directors. "Shot on medium format film, 85mm equivalent, in the style of Paolo Roversi" communicates texture, compression, and aesthetic preference simultaneously. Midjourney has substantial knowledge of fashion photographers and responds accurately to them.

Example Prompt Structures

Quiet luxury editorial

High-energy campaign

Studio editorial

Parameters That Matter for Fashion

--ar (aspect ratio)

Always specify. For Instagram 4:5, use --ar 4:5. For editorial portrait, --ar 3:4. For landscape and OOH, --ar 16:9 or --ar 3:2. The aspect ratio affects composition significantly, Midjourney will adapt the framing to fill the specified dimensions.

--s (stylisation)

Controls how strongly Midjourney applies its own aesthetic judgment. Higher values (700-1000) produce more striking, opinionated images. Lower values (100-400) produce more literal interpretations of your prompt. For editorial fashion, 650-850 tends to produce the most interesting results.

--v (version)

Midjourney v6 is the current standard for fashion work. It has significantly better understanding of photographic lighting, garment rendering, and compositional nuance than earlier versions.

--no (negative prompts)

Use --no to exclude elements that Midjourney tends to add by default. Common exclusions for fashion: --no text, watermark, logo, oversaturated, cartoon, illustration. For a specific aesthetic: --no bright colors, commercial photography, smiling.

The Vocabulary That Fashion Teams Miss

Midjourney's training data includes a large body of fashion photography, which means it has internalized the vocabulary of the industry. Terms that experienced creative directors use naturally produce significantly different results than generic descriptors.

Useful lighting vocabulary: chiaroscuro, Rembrandt lighting, butterfly lighting, practical light source, rim lighting, contre-jour, available light, overcast flat light, dappled light through foliage.

Useful compositional vocabulary: negative space, rule of thirds broken, dead centre, environmental compression, Dutch angle, low horizon, extreme close-up, mid-shot with room.

Useful photographic vocabulary: film grain, lens flare, shallow depth of field, hyperfocal distance, motion blur, shutter drag, overexposed highlights, crushed blacks.

From Brief to Prompt: The Translation Process

The most efficient way to work with Midjourney for fashion is not to write prompts from scratch. It's to write the campaign brief first, the full set of creative decisions, and then translate that brief into prompt syntax.

This approach produces better results for two reasons. First, the creative decisions are made in the right context, when you're thinking about the campaign holistically, rather than under the prompt-writing pressure of trying to produce an image immediately. Second, the same brief can be translated into prompts for multiple models simultaneously, giving you options across Midjourney, Flux Pro, Stable Diffusion, and others without repeating the creative work. Encoding your brand DNA into the brief ensures every model output stays visually consistent with your creative signature.

AI creative direction platforms like the Essenzi Creative Engine handle this translation automatically, taking a structured brief and generating model-specific prompts for 26 different AI tools, with the vocabulary and parameters appropriate to each.