This report describes an initial step in the process of determining whether color vision changes might form the basis for testing to assist in diagnosis of digoxin toxictty. The research questions concerned the types of color vision deficiencies found in people with elevated digoxin levels, the types of color vision tests that could help identify color vision changes in these people, and the relationship between serum digoxin level and subjects' responses to color vision tests. Three groups of subjects with a minimum of one week on maintenance digoxin therapy were tested. Two groups who were free of selected known causes of abnormal color vision participated in a single test session: clinic women (N=19) and hospitalized women (N=30). A third group (N=10) was initially tested at a time of elevated serum digoxin levels (≤2.5 ng/ml) and retested at therapeutic levels. The Farnsworth-Munsell 100-Hue Test was performed on the clinic women. All subjects were tested with the Ishihara; the Hardy, Rand, and Rittler (HRR) plates; and the Farnsworth Panel D-15. The 49 women who had a single test session demonstrated a positive relationship between digoxin level and failing the Ishihara (p < .05). On retest at therapeutic levels, the group with initial high digoxin levels had a significant reduction in the number of Ishihara (p .005) and HRR (p < .005) errors. The Panel D-15 lacked sensitivity. Most subjects with high digoxin levels would not have been able to perform a reliable 100-Hue test. Red-green deficiency was the most common defect.