A wooden shape sorter box with a round, square and triangular hole cut in its lid, and the three matching wooden blocks lying beside it on a cream linen cloth.

What age is a shape sorter for?

A circle drops in at 12 to 15 months, a square and a triangle through the second year, and turning the shape before pushing at about two. The last step is the interesting one.

Every month figure here is a range, and a wide one. A square at 15 months and a square at 22 are both fine.

If you are really asking about puzzles, that is a different job for the hand, covered separately in when do toddlers start doing puzzles. A knob puzzle is flat matching with a handle built in. A sorter asks for more.

The triangle is the hard part

In 2007, Helena Örnkloo and Claes von Hofsten published a study in Developmental Psychology. They gave toddlers aged 14 to 26 months five blocks to post through holes cut to fit them: circular, square, rectangular, ellipsoid and triangular. Each block fit one hole, and only if it was turned the right way.

What they watched was the hand on the way there. Did the child turn the block to match the hole before it arrived, or arrive first and then shove?

At 14 and 18 months the answer was shove. Turning the block before arrival was essentially absent. From 22 months the children turned it in advance and solved the problem consistently. The 22-month-olds managed that for every shape except the square and the ellipsoid. By 26 months they did it for all five.

So the toddler jamming the triangle in sideways and getting cross is doing the only thing available to them. At 18 months, trial and error is the whole toolkit.

Worth being precise about what this shows. One lab, age groups four months apart, a box built for the study. "From 22 months" means the youngest group that managed it, not a switch that flips on a birthday. It does not prove that sorters teach the skill, and we will not pretend otherwise.

A wooden shape sorter box with a round, square and triangular hole cut in its lid, and the three matching wooden blocks lying beside it on a cream linen cloth.
The circle has no wrong way round, which is why it goes in first. The triangle can be held at any angle and only one of them works, and that difference is most of a year.

Which shapes go in first

The same paper gathers up older work on which shapes go in when, which is the ladder most parents are actually asking about. We have not read the originals, so take these as reported.

  • Circle first. Gesell and Thompson (1934) and Bayley (1969) both put a ball into a round hole at about 12 and a half months. The circle is the easy one because there is no wrong way round.
  • The square comes about seven months later. Those same scales put a cube into a square hole at about 19 and a half months. Geerts and colleagues (2003) found most 14-month-olds could already do it. About five months of disagreement between researchers on a near-identical task tells you how much this varies.
  • Triangle last. In pilot data from Achard and von Hofsten (2000), the triangular block went in on 2 percent of tries at 12 months, 22 percent at 16, 60 percent at 20 and 69 percent at 24. The cylinder was already at 74 percent at 12 months.

Even at two, then, the triangle failed about a third of the time. A 20-month-old who gets it in every other try is doing exactly what 20-month-olds do.

The gap between those two shapes is what separates the shelves. For 1 year olds the job is posting and pulling back out, one or two easy holes. For 2 year olds the turning starts, and a fuller set of shapes earns its place.

Five seconds, then hand it over lined up

The same study timed the attempts, and one detail changes what you do at the table. At every age, four out of five successes came in the first five seconds of a try. The 22-month-olds kept at it for around 16 seconds on average, with little to show for it.

In plain terms: if it has not gone in quickly, it is not going to. The extra pushing is mostly effort, and it usually ends with the block on the floor.

  • Hand it over already lined up. Turn the triangle so its point matches the hole, then pass it. The child still does the posting, which is the part they care about.
  • Do not turn it in their hand. Reaching over to rotate the block takes the whole problem away. Reset and pass it again instead.
  • Count to five, then rescue. Five seconds of trying is real work. Twenty seconds of jamming is a child heading for tears, and nothing in the data says the twentieth second teaches more than the fifth.
  • Leave the wrong hole alone. If the square is being pressed against the circle, the child is finding that out for themselves, and the finding is the point. The moment you say "other one", it becomes yours.
If a child has a difficulty and the teacher interferes to show how to deal with it, the child will leave the teacher with the work and go away. The interest of the child was not in the mere task, but in conquering that difficulty.

Maria MontessoriThe Absorbent Mind (Theosophical Publishing House, Adyar, 1949), Chapter XXVII, "The Montessori Teacher," p. 399

What counts as playing with it

Most of what a one-year-old does with a shape sorter looks nothing like sorting. That is fine. It is all use.

  • The lid comes off and everything gets dumped. Legitimate. One hand holding the box while the other pulls the lid is real two-handed work, and at 12 months it is the main event. I think the dumping is the most underrated thing the toy does.
  • Posting through the open top. A circle dropped through the big gap where the lid used to be still went in somewhere. Let it count.
  • Carrying the pieces to another room. Also use. The shapes are good objects on their own, solid and easy to grip.
  • One hole, one shape, over and over. If the circle goes in thirty times and nothing else gets touched, the child has found their level and is drilling it. Do not upgrade them.

Which is why the first sorter should be simple. One hole, a circle-only box, or a drawer-style box where the piece goes in and comes straight back out. A twelve-shape sorter at 13 months is a box of eleven wrong answers.

The simple posting toys live on our fine motor skills shelf. Matching shapes to holes starts on the math and logic shelf from around two.

A wooden shape sorter with its lid set aside and its wooden blocks scattered across a low wooden table in soft daylight.
Lid off, blocks everywhere, nothing posted. At one this is the whole game and it counts. The same scene at two and a half, untouched for a week, means the box is ready for the cupboard.

How to tell it is too hard

Two things look the same from across the room and mean opposite things.

Interest plus frustration is the right level. The child keeps coming back to the box, gets the circle, fights the square, shouts, and comes back after lunch. Hand the square over lined up and leave the rest alone.

No interest after a week is the tell that it is genuinely too hard. Frustration makes noise. This makes none. The box sits there, the pieces get carried off, the lid gets used as a plate, and nobody tries to post anything. Put it in the cupboard and bring it back in six weeks.

If your child is well past two and a half, never posts anything into anything, and it worries you, ask your pediatrician rather than a blog. Charts make this look tidier than it is, and a shape sorter is a poor screening tool.

What this changes at home

Start with one hole, or a circle-only box, at around a year. Add the square when the circle is boring, and the triangle when the square goes in more often than not. Expect the turning-before-pushing step at roughly two, with the square and the egg-shaped block the last to fall.

At the table, line the piece up, pass it, and keep your hands in your lap. The child who spends five seconds on it and then looks up at you has done the work. The child who is twenty seconds in has stopped learning and started grinding. Reset and pass it again.

The sorter you already own is almost certainly fine. The change that matters costs nothing: who does the turning. For the next year that should be the child, wrong angle and all.

Where this came from

  • Örnkloo, H., & von Hofsten, C. (2007). Fitting objects into holes: On the development of spatial cognition skills. Developmental Psychology, 43(2), 404-416. One lab, age groups four months apart and a box built for the study; it shows when turning-in-advance appears, not that a sorter teaches it.
  • The circle-first ladder and the success rates by age come from Gesell and Thompson (1934), Bayley (1969), Geerts et al. (2003) and the Achard and von Hofsten (2000) pilot data, all as reported in Örnkloo and von Hofsten (2007) above. We have not read the originals, the pilot figures come from small groups, and the sources disagree with each other by months on the square.
  • The advice on handing the piece over lined up, counting to five, and reading a week of no interest as the tell rests on the timing data in the same paper plus our own judgment. No study has tested those habits at home, and every month figure in the post is a range, not a deadline.

Sarah G.

Senior Editor & Teacher

We read the research before we write, we say so when it is thin, and we do not dress a marketing claim up as a finding. Where the evidence runs out, the post says where it runs out.

Found something here that is wrong, or out of date? Tell us and we will change it.