Grapevine downy mildew (Plasmopara viticola) is a major disease of European cultivated grapevine (Vitis vinifera L.) against which a large amount of synthetic pesticides are used. Developing microbial biocontrol of P. viticola could reduce the use of pesticides in viticulture and preserve human and environmental health. To achieve this goal, the ecological interactions that develop during infection between P. viticola and the vine foliar microbiome need to be explored. Here, we present metabarcoding datasets describing the bacterial and fungal communities of pairs of symptomatic and asymptomatic leaf samples collected during downy mildew epidemics in three major wine-producing regions of France. Fungal and bacterial communities were sequenced on a MiSeq Illumina platform, and the abundance of the oomycete P. viticola was quantified using qPCR. We provide the raw metabarcoding datasets, the amplicon sequence variant tables obtained after bioinformatic processing, the metadata describing sampling sites and tissue health conditions, and the code used for bioinformatic analysis. These datasets will enable microbiome comparison within pairs of symptomatic and asymptomatic samples collected at the same time on the same leaf. Such a comparison could help describe the ecological interactions between P. viticola and the grapevine foliar microbiome.