Tree diebacks are complex and multi-factorial diseases with suspected biotic and abiotic components. Microbiome effects on plant health are challenging to assess due to the complexity of fungal and bacterial communities. Grapevine wood dieback is the main threat to sustainable production worldwide and no causality with microbial species has been established. Here, we aimed to test the hypothesis that grapevine esca disease progression has reproducible drivers in the fungal species community. For this, we analyzed a set of 21 vineyards planted simultaneously with a single susceptible cultivar to provide replication at the landscape scale. We sampled a total of 496 plants across vineyards in two different years to perform deep amplicon sequencing analyses of the fungal communities inhabiting grapevine trunks. The communities were highly diverse with a total of 4,129 amplified sequence variants assigned to 697 distinct species. Individual plants varied in fungal community composition depending on the year of sampling, vineyard location, and disease status. However, we detect no specific fungal species driving symptom development across the vineyards contrary to long-standing expectations. Our study shows how landscape-scale replicated field surveys allow for powerful hypothesis-testing for complex dieback disease drivers and prioritize future research towards additional factors.