An Experiment in Painting with Clay

After returning from Geeli Mitti in November, I knew I wanted to paint one wall of my apartment in Delhi with clay paint. The idea stayed with me for weeks. The wall had already suffered damage during the monsoons and needed maintenance anyway. More importantly, the source of dampness—something that returned every monsoon—had finally been resolved. That gave me the confidence to try something different.

The original condition of the wall in my apartment in Delhi.

I had attended the Intensive Natural Building Workshop at Geeli Mitti in October–November 2025. During the workshop, we learned to prepare clay building materials and natural paints, among many other things that deserve stories of their own. What stayed with me was how intuitive the process felt—like cooking or play. Working with natural materials doesn’t feel mechanical; it feels personal. You touch it, sense it, mould it. It’s less about instructions and more about listening.

At Gilli Mitti.

Months earlier, in July, I had collected soil from my family’s farmland in Kaushambi, Prayagraj, and stored it in a bottle. The original intention was to test the soil and understand what clay building techniques could be used for my parents’ farmhouse someday, though I was never fully sure if that would happen. When I tested the soil, it turned out to be mostly clay with very little sand. Its natural colour was a deep brown.

I wasn’t certain about the colour I wanted for the wall, so I bought white clay (kaolin) and marble dust online to experiment alongside the site soil. Before starting, I revisited a few resources—my notes from Geeli Mitti, along with videos by Geeli Mitti and SunDogBuilder—to better understand proportions and process.

Clay paint, at its core, is simple. It needs a base and binder, an aggregate, and a pigment. In my case, the site soil acted as both base/binder and pigment. Wheat paste was added as an additional binder to prevent dusting. Marble dust became the aggregate (optional, but helpful for texture and colour control, especially given my limited material quantity). Pigments can be anything natural—flowers, leaves, turmeric, beetroot—or none at all, if you choose to stay with the clay’s own colour. Wanting to avoid synthetic pigments, I extracted colour from raw turmeric and beetroot. The general proportion I followed was one part clay, one part wheat paste, one part aggregate, and water as required to achieve a paint-like consistency.

Before committing to the final paint, I needed to make samples—to test colours, textures, and how well the paint would adhere to the existing painted and plastered wall. The process began with sieving the site soil using an old tea strainer. Then came the wheat paste, made by mixing refined flour (1 cup) with room-temperature water (1½ cups) until smooth, then slowly pouring it into hot water (2 cups) over a low flame, continuously stirring to form a thick, consistent paste.

I made several samples: white, beetroot, turmeric, and combinations of site soil and kaolin. All of these included marble dust, except one sample made with just site soil and wheat paste. I tested proportions freely, choosing colours I felt comfortable living with in my bedroom.

Sampling on the wall.

The first layer was discouraging. Some samples were too watery, others too dry. The paint didn’t spread evenly, and it felt like the experiment had failed. I left it to dry for a few hours and tried again. The second layer came out much better—more even, more consistent—though areas where the first layer hadn’t fully dried rubbed off slightly. Still, it felt promising.

Applying first layer.

After 24 hours, the samples told their story. The beetroot pigment had almost disappeared. All colours had lightened significantly. The darkest and most stable result was the site soil mixed only with wheat paste.

I invited a few friends to help, hoping to finish the wall in a day. Their presence added to the fun. We finalised the colour and began preparing the paint in small batches, gradually increasing the quantity while keeping proportions consistent. The wall measured 11 feet by 9 feet, and we eventually used about 7–8 litres of paint.

Before painting, the wall was cleaned with a metal wire brush. While removing the sample patches, we noticed something interesting: the site soil and wheat paste sample refused to come off, almost as if it had fused with the existing layers of paint, putty, and cement plaster.

After applying first coat.

The first coat was applied with a brush. We initially planned to let it dry for a day before adding another layer, but a few hours later, one friend suggested using a roller. The second coat was done with a roller, and hands were used to smooth edges and uneven spots. Painting with bare hands felt natural—and easier.

By the end of the second layer, the old yellow paint underneath was no longer visible. The wall looked complete. Almost all the paint was used; only a few hundred grams remained. The next day, as expected, the colour lightened further. This was exactly why sampling mattered.

After applying second coat.

The final result was a soft beige—textured and calm. It didn’t look like clay paint in the way people imagine. It looked like a regular painted wall, just quieter. Softer. Alive.

What stayed with me most was the feeling. There was no smell of synthetic paint, no harsh chemicals, and the finish felt just as refined. By the end, I found myself wondering if I should paint the entire room the same way. It didn’t feel like a task completed. It felt like time well spent.