This is a very interesting project about transmitting an audio signal over a laser beam. The idea came to me after analyzing the driver circuit of a cheap keychain laser I've disassembled.  This laser diode may be easily modulated in frequency (turning the laser on and off rapidly) thus, the input signal amplitude is transformed in small frequency variations over a central "carrier" frequency, this modulating technique is called FPM -Frequency Pulse Modulation-

An example of the general arrangement of the laser transmitter TX and receiver RX is given in the following picture. In this case the signal may be bounced over a reflecting surface (retrocube for example) or just making contact with another station identical to this one.

It's important to note that both modules (TX & RX) are independant from each other, thus preventing interferences.

The transmitting module TX circuit is built around a NE555 timer controlled on its pin 5. Surely there are other integrated circuits more appropiated for FPM, but in this case I used a 555 because they are cheap and come handy.

The receiver module RX is built using a NE567 tone deocoder and a TDA7052 general purpouse integrated audio amplifier. Again, the NE567 it's not properly a dedicated PLL (like NE565) but it was what I had at the moment and worked pretty well despite its limitations.
There's a lot of room for sensor developement, not only by replacing the phototransistor by a pin diode but also using a operational amplifiers at the input stage (like a CA3140).

The MRD300 (Motorola) is a 50mA / 250mW NPN phototransistor. See detailed photos below, note pin identification (collector-base-emitter). Equivalent units could be MTD6040, L14G2, BPW14, OP804, BPX43-3, TIL81, TIL78, also a BPW40 may be used. In fact any photo- transistor o diode will do (system response may vary accordingly!)

Finally it's quite important to have something to concentrate the incoming signal, it may be a concave mirror, a lens or any kind of telescope.
As i had a 60mm diameter lens from an old Xerox machine around, I decided to use it as the main receiving surface. Also, as a laser beam is so directional, proper alignement is critical, so nothing better than improvise a "scope" for easy align the Rx with the TX module.
The scope is depicted below.

Actual photos of the scope and phototransistor attachment.