Somewhere between seven and twelve, depending on the child and — more importantly — on what you start them with. The age question gets asked a lot; the starting-point question matters more, and getting it wrong is the most common way children decide they are "not a coding person" before they have really tried.

What children can handle, roughly by age

These are tendencies, not rules. Plenty of nine-year-olds are ready for things this list puts at twelve.

  • Ages 6–8 — sequencing and logic through block-based tools and unplugged games. The concept being taught is "instructions happen in order", not programming.
  • Ages 9–11 — loops, conditions and variables, still mostly visual. This is where most children can genuinely build something they designed.
  • Ages 12–14 — a real text-based language. Typing accuracy and tolerance for syntax errors are the actual prerequisites, not mathematical ability.
  • Ages 15+ — data, libraries and projects with real complexity. Python and pandas sit comfortably here.

The mistake that puts children off

Starting with a language whose errors are cryptic. A ten-year-old who writes a missing bracket and gets a wall of red text does not learn about syntax — they learn that computers are hostile and they are bad at this. That impression can last years.

This is why we start younger students on Small Basic rather than jumping straight to Python. It is a real language with real syntax and real errors, but a fraction of the ceremony, and something appears on screen within the first ten minutes. Students move to Python afterwards without unlearning anything.

Maths ability is not the prerequisite people think

Parents frequently tell us their child is not strong at maths and therefore probably cannot code. For everything a child will do in their first two years of programming, this is untrue. Early coding is about sequencing, decomposition and patience — closer to writing instructions than to algebra.

Maths becomes relevant later, in specific areas: graphics, simulation, data science, machine learning. It is not a gate on entry, and treating it as one excludes children who would enjoy this a great deal.

Signs a child is ready

More useful than age: can they follow a multi-step instruction without losing the thread? Do they persist for a few minutes when something does not work, rather than immediately asking for help? Do they take things apart — physically or in games — to see how they function?

A child who says "I want to make a game" is ready. A child who has been told they should learn to code because it is good for their future is a harder start, and usually needs a project they chose themselves before anything sticks.

How much screen time is this really?

A fair question, and worth separating from general screen time. There is a meaningful difference between consuming and producing. An hour spent building something that did not exist before is not the same hour as watching short videos, even though both look identical from across the room.

One session a week, with something to show at the end of it, is a reasonable amount for most children.

What good online learning for kids looks like here

Small groups, because a child stuck on a bracket needs a teacher who notices within thirty seconds, not at the end of the lesson. Live rather than recorded, for the same reason. And projects the child picked, because the difference in effort between an assigned exercise and their own idea is enormous at this age.

Our coding courses run in groups capped at five for exactly that reason. Fun with Coding suits first-timers from around ten; Learn Python suits students ready for a text-based language, usually twelve and up; and AI for Kids covers how these tools work and how to use them honestly — increasingly the conversation parents actually need to have.

Written by the Learning Bubble teaching team. If something here does not match your situation, ask us — we would rather give you a straight answer than have you guess.