Here's an idea that might be a bit complex, but it shouldn't be too hard
to implement:

Have an oscilator which produces a square wave a a few 100 MHz. Feed it
through a BPF to shape it into a sine wave and feed it to a high speed
laser diode driver. Place a reflector at the location to which you are
trying to find the range. Have a PIN photodiode receiver with another BPF
and amplifier chain (this receiver is located back near the laser). The
output of this receiver is sent thru a fast schmitt trigger and then to a
fast counter. A second fast counter counts the output of the oscilator
directly. The PIC (yes, there is a pic :) ) triggers the whole system by
feeding the oscilator to both the laser and one of the fast counters, and
clearing both counters. It then waits the maximum ranging time of flight,
and turns off the OSC. THe difference between the numbers on the two
counters represents the number of cycles that occured during the TOF.

I think that this is superios to the single counter method becvause in
that case, something has to detect the reflected signal very quickly and
stop the count. I think that it can be very difficult to sense a signal
reliably in only a few cycles. If a false detect occurs  with the single
counter, the result will be way off. With the two counters, it will just
be off by the numbver of spurious cycles.

Sean