A new murine model of anxiety, namely the mirror chamber test, is based on the assumption that, like many species, mice show approach-avoidance behavior when they are confronted by a mirror. It has been suggested that the mirror chamber is a specific and a quantitatively/qualitatively different measure of anxiety than that implicated in other behavioral models such as the elevated plus-maze and the head-dipping assays. The aim of the present study was to investigate whether there was indeed a specific mirror effect by replacing the mirrors in the chamber by either white or dark-gray tiles. Balb/c, DBA-derived, and C57 BL/6j mice were tested under these three experimental conditions. The results indicated that Balb/c and DBA-derived mice avoided the mirror, white and gray chambers similarly while C57 BL/6j mice entered more readily and spent more time in the gray chamber than in the mirrored or white chambers. Thus, depending on the strain of mouse studied, a brightness or a position effect in the chamber could explain the avoidance behavior observed. These results suggest that there is no need to invoke a specific mirror effect.