This experimental study examined the validity of 3 underlying mechanisms that have been proposed in the research literature to explain the anxiety-reducing effect of coloring mandalas: distraction, structure, and centering. Four conditions were tested: (a) freely coloring a mandala, (b) coloring a mandala with preset colors, (c) freely drawing a mandala within a preset circle, and (d) coloring a rectangular grid with preset colors. Anxiety was induced in the participants and then measured by self-report and pulse rate. The anxiety-reducing effect of the intervention task was found to be significant in all 3 mandala conditions as compared to the control condition. The centering explanation was supported by the physiological data but not the self-reported data. The structure explanation was not supported by either source of data. The distraction explanation could not be verified because the manipulation of distraction was not successful. Findings from this study support the use of mandalas as an anxiety-reduction tool in art therapy, however the validity of the distraction effect needs further investigation.