(1) The object of the search described in this publication was to find “inhibitors of virus increase”, not “inhibitors of infection”. (2) The investigations were performed with the following plants and viruses: tobacco species with tobacco mosaic virus (TMV); the disease causes leaf damage; sugar beet with virus yellows; the disease decreases root production; French bean and broad bean with Phaseolus viruses 1 and 2; the disease reduces seed production. The cultivation method, the age of the plants tested, the virus, its purification and properties, and the techniques applied to infect the plants are briefly described. The virus influence was investigated by means of direct methods, such as serology, and indirect ones, such as a reducing sugar test according to Van Duuren (1955), a reso blue test, (Münster (1955), Schuster (1956)), to determine the callus production in the phloem (these two for sugar beet only), and a test with 2,3,5-triphenyl tetrazolium chloride (TTC) (see F. E. Smith, 1951, and Beal et al., 1955). All the indirect tests, as well as the various tests with Nicotiana glutinosa, proved to be of little value (Tables n, in and iv). (3) The chemicals were applied via the leaves (spraying, preferably as a 0.1 % solution, at a rate of 100 ml per square metre) or via the roots (dipping in a 0.1 % solution, or nutrient technique, 0.001-0.01 %w added to 150-200 ml Knopp’s solution). Table v lists the chemicals with antivirus properties described in the literature. Some of the most important are discussed separately (thiouracil, 8-azaguanine, malachite green, growth regulators and inorganic salts). On the basis of these data several types of chemicals were investigated. The nitroso hydroxyaryl compounds displayed such an activity that closer examination was justified. The antivirus activity of a chemical was expressed as the percentage increase in incubation time; for this purpose the exact moment at which the first visible symptoms of the disease appear must be determined. p-Nitrosophenol, l-nitroso-2-naphthol, and the copper sulphate complex of o-nitroso-p-chlorophenol proved very attractive, being not too phytotoxic (Tables vi, vxi and vm); the potassium salt of l-nitroso-2-naphthoI-3,6-disulphonic acid showed good activity, but also some phytotoxicity. Thiouracil and its mixtures with p-nitrosophenol and l-nitroso-2-naphthoI displayed a particularly good activity on tobacco without being phytotoxic in the concentrations used (0.05 % + 0.05 %; Table vm). (4) The four nitroso compounds mentioned were used in field experiments. In some tentative trials on French bean and broad bean the antivirus properties of p-nitrosophenol and l-nitroso-2-naphthol were again observed (Table x). Most experiments, however, were performed on sugar beet. Indications of antivirus activity were obtained in all cases. Taking all the results obtained in 4 years (56 experiments) together, an average improvement of 7 % ± 1 as compared with untreated plants was obtained for l-nitroso-2-naphthol, mostly applied as its potassium salt (Graph 1). The other nitroso derivatives mentioned did not show any significant antivirus activity. The nitroso-thiouracil mixtures had a very reliable activity on tobacco, but on other plants their phytotoxicity proved to be too high (Tables xm and xtv). (5) It was demonstrated that these nitroso compounds only have a good effect when they are applied in a 0.1 % concentration. Two or at most three sprayings are permissible. A higher concentration or more frequent spraying produces phytotoxic effects, often to such an extent that the resulting loss exceeds the improvement brought about by the antivirus compound. The influence of the chemical could be demonstrated with serological tests, but not with the other tests mentioned under (2). The chemical itself is decomposed in the plant. It may be that it reacts with uracil, just as thiouracil does, or with any other indispensable virus component available in the plant cells. It can thus be expected that there is not enough uracil left for the virus synthesis in the cells. As soon as the nitroso compound is entirely decomposed the uracil formed in the plant again becomes available for virus synthesis and the plant is attacked by the disease after all; only the incubation time has increased considerably. If this hypothesis should be true the nitroso compound would act as an antivirus factor. On the other hand, such a hypothesis would also explain the fact that the phytotoxic and antivirus activities of these compounds are almost the same, as the plant itself needs uracil as well.