Mike
I , like Sergey, use m. to reference variables as a general rule. I had heard many times that there was a slight performance hit, but I have found it to be better to avoid problems. This thread got me curious, so I wrote a little test program to check.

m.test1 = 0
m.test2 = 0
m.test3 = 0
m.test4 = 0
m.test5 = 0
m.test6 = 0

m.seconds1 = seconds()

for m.xx = 1 to 500000
	m.test1 =  m.test1+1
	m.test2 =  m.test2+1
	m.test3 =  m.test2 +1
	m.test4 =  m.test4+1
	m.test5 =  m.test5+1
	m.test6 =  m.test6 +1
next
?strtofile("using m. for all variables in 1 to  500000 loop"+chr(13)+chr(10),"test.txt",.t.)
?strtofile(alltrim(str(seconds()-m.seconds1,10,3))+" Seconds"+chr(13)+chr(10),"test.txt",.t.)
test1 = 0
test2 = 0
test3 = 0
test4 = 0
test5 = 0
test6 = 0
seconds1 = seconds()

for xx = 1 to 500000
	test1 =   test1+1
	test2 =   test2+1
	test3 =   test2+1
	test4 =   test4+1
	test5 =   test5+1
	test6 =   test6 +1
next
?strtofile("NOT using m. for all variables in 1 to  500000 loop"+chr(13)+chr(10),"test.txt",.t.)
?strtofile(alltrim(str(seconds()-m.seconds1,10,3))+" Seconds"+chr(13)+chr(10),"test.txt",.t.)

Results using P4 1.8Ghz 256MB VFP 6.0 SP5

Using m. for all variables in 1 to  500000 loop
1.892 Seconds
NOT using m. for all variables in 1 to  500000 loop
1.803 Seconds

Using m. for all variables in 1 to  1000000 loop
3.775 Seconds
NOT using m. for all variables in 1 to  1000000 loop
3.566 Seconds

Using m. for all variables in 1 to  1500000 loop
5.668 Seconds
NOT using m. for all variables in 1 to  1500000 loop
5.298 Seconds

To me this difference in performance is neglible compared to the benefits that I have found in readablilty of code and preventing ambiguity and confusion especially when looking at code some years later.