Why not use random.randint
instead, so you don't have to use list.index
to find the list later:
from random import randint
# your list of lists
l = [[1,2,3],[4,5,6],[7,8,9]]
# choose a valid *index* into l, at random
index = randint(0,len(l) - 1)
# use the randomly chosen index to get a reference to the list
choice = l[index]
# write your conditionals which handle different choices
if index == 1:
print 'first list'
elif index == 2:
print 'second list'
...
This will be more efficient than using random.choice
and then list.index
each time you make a choice.