There seems to be a significant regression in performance when using @property in Python 2.x compared to 3.x.

Test code:

class A:
    def __init__(self, x, y):
        self.x = x
        self.y = y
    @property
    def y(self):
        return self._y
    @y.setter
    def y(self, value):
        self._y = value

from timeit import timeit
a = A(1, 2)
timeit('a.x', 'from __main__ import a')
timeit('a.y', 'from __main__ import a')


On my machine (Fedora Linux, x64) I get the following timings:

2.7.8:
a.x : 0.05482792854309082
a.y : 0.05585598945617676

3.4.1:
a.x : 0.06391137995524332
a.y : 0.31193224899470806

I.e. The performace of using a property vs a ordinary member is more or less the same in 2.7, while it incurs a 5x penalty in 3.4.