In 1993, control failures were reported on grapevine in northern Italy under severe downy mildew pressure after postinfection application of cymoxanil in mixtures with copper or mancozeb. A monitoring survey was started immediately in Piedmont (north‐western Italy) in order to determine the sensitivity of populations of Plasmopara viticola to cymoxanil from those vineyards where the fungicide was not controlling the disease satisfactorily. In 1994 and 1995, monitoring surveys were extended to north‐eastern Italy, where cymoxanil mixtures were not performing as well as in the past. Sampled populations were tested on detached leaf discs and on whole potted plants under controlled conditions. In 1993, 12 populations, sampled in Piedmont, showed MIC values (minimum inhibitory concentration) varying from 10 to more than 100 mg L−1 cymoxanil. With a baseline reference population having a MIC value of 3 mg L−1, resistance factors ranged from 3 to more than 30. In 1994, 17 populations out of 27 sampled in Trentino (north‐eastern Italy) showed MIC values of 100 mg L−1 or higher and in 1995, 32 populations out of 38 showed the same behaviour. In similar experiments, the MIC values of populations from nontreated plots were between 3 and 10 mg L−1. In whole potted plant tests, populations with MIC values higher than 200 mg L−1 in a leaf disc test were not controlled by 500 mg L−1 of cymoxanil. The results of our study suggest that resistance to cymoxanil in P. viticola may contribute to a lack of disease control in Italian vineyards.