Quick idea

Pythagoras' theorem says for a right triangle with legs a and b and hypotenuse c: a² + b² = c². A Pythagorean triple is a set of whole numbers (a, b, c) that satisfy this (like 3, 4, 5). The "hardest-known Pythagoras problem" most people mean is the perfect cuboid problem: can you find a rectangular box with whole-number edges a, b, c so that every face diagonal and the space diagonal are all whole numbers too?

What the problem asks (in equations)

We want integers a, b, c so that these six values are integers:

• edges: a, b, c
• face diagonals: sqrt(a² + b²), sqrt(a² + c²), sqrt(b² + c²)
• space (body) diagonal: sqrt(a² + b² + c²)

If the three face diagonals and the edges are all integers, we call such a box an "Euler brick." If the space diagonal is also an integer, it's called a "perfect cuboid." No one has found a perfect cuboid, and no one has proved that one cannot exist — that's why it's an open (and famous) problem.

Examples so you see what's possible

We know many Pythagorean triples and we can make boxes where all three face diagonals are integers (Euler bricks). A well-known Euler brick is:

a = 44, b = 117, c = 240

• sqrt(44² + 117²) = 125
• sqrt(44² + 240²) = 244
• sqrt(117² + 240²) = 267

But the space diagonal sqrt(44² + 117² + 240²) is NOT an integer. So this is an Euler brick, but not a perfect cuboid.

Why is it hard?

1. It combines several Pythagorean equations at once. You must solve the system
   a² + b² = d²,
   a² + c² = e²,
   b² + c² = f²,
   and also a² + b² + c² = g².
   Each is a Diophantine equation (integer solutions only). Getting them all at once is very restrictive.
2. Although Pythagorean triples have a complete formula (for primitive ones: a = m² - n², b = 2mn, c = m² + n²), combining three such parameterizations so they match up to share edges is messy and leads to hard algebraic conditions.
3. People have used deep number theory and massive computer searches but still neither a perfect cuboid nor a proof of impossibility has been found. That combination of simple statement + deep difficulty is what makes it famous.

What mathematicians know (briefly)

• Euler bricks (integer edges and face diagonals) exist — many examples are known.
• No perfect cuboid (with integer space diagonal as well) has been found despite large computer searches.
• There are partial theoretical constraints: certain congruences and divisibility properties must hold for any potential perfect cuboid, so not every choice of a, b, c could work. Those rules reduce the search but haven't ruled out all possibilities.

How you could explore this yourself

1. Start small: list Pythagorean triples (3,4,5), (5,12,13), (8,15,17), etc. Use the m,n formula to generate more.
2. Try to combine triples to make two face diagonals integer. For example, pick (a,b,d) and (a,c,e) both triples so a is shared and both pairs give integer diagonals.
3. Write a small program (Python works great) to search for Euler bricks by checking many pairs (a,b,c) and testing whether the three face diagonals are integers. Then check the space diagonal too.
4. Read about more advanced techniques once you’re comfortable — number theory papers discuss modular constraints and other filters that speed up the search.

Summary

The perfect cuboid problem is a simple-to-state, decades-old unsolved problem: does a rectangular box exist with integer edges, face diagonals, and space diagonal? We know many boxes where the face diagonals are integers (Euler bricks), but nobody has found one with the space diagonal integer, and no proof says one cannot exist. It’s a great example of how an easy-to-understand question can lead to deep mathematics.

Want a next step? Try generating Pythagorean triples and then write a small search that looks for Euler bricks — it’s a fun coding + math project and gives real insight into why the problem is so challenging.