The instinct to fill the page
A worksheet with eight questions on it looks unfinished. There is white space at the bottom, it takes thirty seconds to print and it feels like it cannot be enough. So most worksheets get filled — thirty, forty, fifty problems, because that is what fits.
That instinct is worth resisting, because sheet length is not a neutral formatting decision. It changes what the child does, how accurate they are by the end, and what you can legitimately conclude from the marks.
What happens in the last row
Watch a child work through forty similar problems and the pattern is consistent. The first ten are careful. The middle twenty get faster and the handwriting deteriorates. The last ten are raced, because the end is visible and the task has become something to finish rather than something to do.
The errors in that last row are not the same kind of errors as the ones at the top. Early mistakes tell you about understanding. Late mistakes tell you about fatigue and motivation, and they are frequently in problems the child could do perfectly well twenty minutes earlier.
This matters because the two get averaged together into a single score. A sheet marked 31 out of 40 looks like a child who mostly understands with some gaps. If eight of the nine errors are in the final ten questions, the real picture is a child who understands it and got tired — a completely different conclusion calling for a completely different response.
Practice and diagnosis want different lengths
It helps to be clear about which of two things a sheet is for, because they do not want the same design.
A diagnostic sheet is trying to find out what a child can do. It should be short — ten to fifteen questions — and every question should test the thing you are checking. If you want to know whether subtraction with exchanging is secure, every problem should require exchanging. Mixing in problems that do not lets a child score respectably while getting every relevant question wrong, which defeats the purpose entirely.
A practice sheet is trying to build fluency in something already understood. Here more repetition genuinely helps, and twenty-five to forty short items is reasonable — but only for recall of facts, not for multi-step methods. Forty long-division problems is not fluency practice, it is an endurance test.
Some numbers to start from
For children of five to six, eight to twelve questions. Attention for focused written work is roughly ten minutes at this age, and a sheet that outlasts it teaches the child that worksheets are something you fail to finish.
For six to eight, fifteen to twenty for most purposes. This is the range where a sheet is long enough to show whether a method is secure and short enough to be finished properly.
For eight and over, twenty to thirty for written method, or up to forty for timed recall of facts — multiplication tables being the obvious case, where the whole point is speed on things already known.
These are starting points, not rules. The better test is empirical: if the last few answers on a returned sheet are consistently worse than the first few, the sheet is too long for that child, whatever the number says.
When ten correct answers is the end of the exercise
If a child gets ten problems right in a row, the eleventh teaches nothing. They have demonstrated the skill. Continuing produces practice at something already mastered while the child's attention drains away, and the remaining questions measure patience rather than ability.
The useful response is to change the difficulty rather than extend the sheet. Raise the number range, introduce the carrying you had switched off, move from horizontal to vertical layout. A short sheet at the right level beats a long one at the wrong level by a margin that is hard to overstate.
This is also the argument for generating sheets rather than reusing printed ones. When the level needs to change, it should take thirty seconds, not a trip to find a different book.
Two shorter sheets beat one long one
If you want thirty questions' worth of practice, two sheets of fifteen on different days will produce better retention than one sheet of thirty, and it is not close. Spacing practice out is one of the most reliably demonstrated findings in how people learn, and it costs nothing to apply.
The same logic works within a single session for younger children: a page combining a short arithmetic block, a few traced words and a small matching exercise holds attention far better than a page of one thing, because each change of activity resets the clock a little.