Dry root rots caused by Fusarium culmorum are an increasing constraint to wheat production in Moroccan semi-arid regions. Wild relatives such as Aegilops tauschii and Triticum dicoccoides represent valuable sources of resistance. Developing a reliable screening method is crucial for identifying promising breeding lines. Six accessions from each wild species were evaluated under controlled conditions using a split-plot design with three replications, where soil type (sterilized vs. natural) was assigned to main plots and inoculum type (organic vs. spore suspension) to subplots. For each accession in each treatment, one pot containing five plants was established per block. Plants were inoculated at the two-leaf stage after vernalization, and disease severity was assessed at flowering. Root and internode infection, spike number, dry biomass, and plant height were measured to integrate both pathological and agronomic responses. Results showed that methodological choices strongly influenced the ability to discriminate between resistant and susceptible lines. Internode assessment was the most sensitive criterion, revealing clear genetic variability both within and between species. Suspension inoculum consistently generated higher and more uniform disease pressure than organic inoculum. At the same time, soil type influenced resistance expression differently across species: in A. tauschii, non-sterilized soil amplified root severity, whereas in T. dicoccoides responses were stable across soil types. Agronomic traits further supported the approach. In A. tauschii, spike number depended on genotype × inoculum and genotype × soil interactions, while in T. dicoccoides it was largely genotype-dependent. Biomass also varied with genotype × inoculum interactions. Plant height exhibited the most complex pattern: in A. tauschii, it was primarily driven by genotype × inoculum interactions, whereas in T. dicoccoides, it was influenced by the combined effects of genotype, soil type, and inoculum. Overall, combining suspension inoculum, non-sterilized soil, and internode evaluation provides a robust screening method for identifying resistant germplasm in wheat improvement programs.