BMV development was studied in well defined physiological conditions, avoiding the use of antibiotics or excision of the roots. For an adequate labelling, to distinguish between host and viral RNAs, different periods of exposure to32 were required, depending upon the post inoculation time of infected plants and corresponding age of healthy control. Within the period of normal outlook of the plants, the rate of32 P incorporation into BMV RNA was much higher than into barley ribosomal RNA in infected tissue or in comparable healthy tissue. Analysis of the replicative structures isolated from infected plants revealed that only the three largest RNAs (A,A,B) had their own replicative intermediates (RI). This material was shown to contain a certain amount of intact viral RNA species, which were stable when treated with heat and formamide, excluding the possibility of hidden breaks.