Comment on: "Electromagnetically induced left-handedness in optically excited four-level atomic media"

We discuss the recent proposal by Thommen and Mandel (Phys. Rev. Lett. {\bf 96}, 053601 (2006)) for electromagnetically induced negative refraction. Although the main conclusion of the paper -- the possibility to achieve negative refraction in an experimentally accessible atomic scheme -- remains valid, we show that the weak-excitation approximation used is invalid in the parameter regime studied and leads to quantitatively incorrect predictions. We show that negative refraction is always accompanied by absorption rather than by gain, and that the maximum value of the refraction-absorption ratio is of order unity.