Control of the metalaxyl resistant strain of Plasmopara halstedii (causal agent of sunflower downy mildew) was studied in greenhouse and field experiments over a three-year period. Of 30 fungicides investigated, only azoxystrobin and fenamidone showed good efficacy as seed treatments for control of systemic symptoms initiated by root infection. Four fungicides displaying lesser disease control (ethaboxam, fluazinam, mancozeb, and zoximide) were tested as co-fungicides with azoxystrobin and fenamidone. None of the fungicides, either used alone or in combination, gave complete control under natural or inoculated field conditions, in contrast to the total control previously experienced with metalaxyl. Disease incidence on all fungicide treatments increased from 2 wk to 7 wk after emergence, suggesting that the fungicide/rates tested were more fungistatic than fungicidal. Fenamidone was the most effective fungicide used singly, and fenamidone combined with either ethaboxam or zoximide was the most effective combination. Greenhouse tests of fungicide seed treatments were not effective at predicting field response of the best treatments, despite several modifications. The greenhouse tests, however, were satisfactory to identify ineffective fungicides. None of the three fungicides likely to be registered as sunflower seed treatments in the United States, namely azoxystrobin, fenamidone, and zoximide, displayed any potential phytotoxicity or seed safety concerns in preliminary laboratory and field emergence tests, intended to run for two years.