A lot of 792 tubers, of which about 20% were infected with the potato spindle tuber virus (PSTV), were subjected to greenhouse index tests, using tomato as a test plant. Tomato seedlings in the cotyledon stage were rubbed with the test material, either foliage or the surfaces of seed pieces cut from test tubers with a mellon-baller. Fourteen days later the test plants were inoculated with a severe strain of PSTV to detect the presence of mild strains. Tubers giving negative results were planted for propogation and those giving positive or doubtful results were planted in an observation plot. All potato plants were inspected visually and those in the observation plot were given an additional tomato test. The results indicate the validity and value of the tomato test for the elimination of PSTV from seed stocks before planting. The tuber test conducted in the winter was less efficient than the foliage test conducted in the spring. A modification of the test procedure is advocated in which two test plants are rubbed from each source, one to be challenged and one not. This increases the accuracy of the test, particularly during periods when symptoms of PSTV are poorly expressed. © 1969 Springer.