Summary.-Ss reported autokinetic movement under 1 of 4 color conditions (red, green, violet, and yellow). The over-all main effect of color on magnitude of reported movement was not significant. The interaction between color and trials was significant. Marone and Coutu (1969) and Reeves, Cogan, and Cogan (1970) studied the effect of color on various reports of autokinetic movement. Results showed that color did not affect autokinetic reports. However, several Ss in the former , study gave no movement responses to a red color stimulus. This was not the case in the Reeves, et al. study. As these results are discrepant, the present study explored the effect of color (yellow, red, violet, and green) on magnitude of reported autokinecic movement. Each of 40 college freshmen with normal color vision were dark-adapted for 30 min. The stimulus source of autokinecic movement was a 7-w light in a box painted flat black. Color was produced by placing either a red (715 nm.), yellow (580 nm.), green (540 nm.), or violet (420 nm.) gel filter over a pinpoint hole in the box. Light passing through each of the filters produced an equivalent amount of luminance (1 X ft-c.) as measured by a McBeth illuminometer placed 8 fc. from the light source. To prevent possible light dispersion, a section of frosted glass was placed over the pinpoint opening and gel filters. S sat in a chair 8 ft. from the light source adjusted to S's eye level. Each S was randomly assigned to be exposed to only one color of light for 5 practice uials and 10 experimental trials each lasting 60 sec., intertrial interval was 15 sec. S's task was to estimate verbally extent of observed autokinetic movement. In addition, S was required to select a strip of paper, from various lengths available, which corresponded closest in length to the extent of movement observed. The lengths of the available paper strips ranged from .5 in. to 12.5 in. in 1-in. increments. ,Verbal estimates of observed autokinecic movement and matching the observed movement to paper strips of corresponding lengths correlated .87 (p < ,001 j. Since this correlation was significant, subsequent analyses concerned only verbal estimates of autokinetic movement. An analysis of variance was performed on reported estimates of magnitude of autokinetic movement by Ss with color as a between-S variable and trials as a within-S variable. Over-all, color was nonsignificant (F = 2.66, df = 3/36,