Why is it harder to hear someone underwater than on air? Because water is denser than air, sound waves travel faster and with more energy in water than air. However, we are worse at hearing in water than in air. Why is this?
To clarify, I was comparing these two:

*

*Having both the sound source and the listener (human) underwater

*Having both the sound source and the listener (human) above water

Supposedly, sound waves are 'better' with denser media, but we humans cannot hear very well underwater.
 A: Impedance mismatch.
The impedance ratio or the admittance ratio (admittance = inverse of impedance) describes how much of a wave is reflected or transmitted at the boundary of two media depending on the frequency. In principle, the ear is an impedance transducer that converts sound waves hitting the eardrum into smaller, more powerful vibrations by means of the auditory ossicles, which act on the cochlea. If the medium acting on the ear is water instead of air, to which the eardrum is optimised, there is an impedance mismatch and the waves are largely reflected instead and lead to only minor vibrations of the eardrum.
A: TL; DR: most of the sound (energy) is reflected from water surface.
If we think about an actual setup of an experiment when we cannot hear somebody while underwater, the person that we want to hear is not in the water: they are speaking in the air and the sound waves have to propagate through the boundary between the air and the water. Water, being denser than air, reflects most of the sound incident on it (like a mirror reflecting EM waves) - only small fraction of the energy of the initial signal penetrates.
The situation is quite different, if we are dealing with sound created underwater, by animals or devices adapted to creating underwater sound.
This could be described as impedance mismatch, and, in this sense, my answer is complimentary to that by @Bulbasaur, who focused on the impedance mismatch between the medium (water) and the ear (adapted to match the acoustic impedance of air.)
