Phelipanche ramosa is a major holoparasitic weed, which became a major agronomic problem for rapeseed crops in France. Weed control in rapeseed does not have a direct impact on this parasitic weed whose half of the reproductive cycle occurs on the root system of its host. Since 1990, broomrape expanded its distribution area starting from the Poitou-Charentes region. Necrosis symptoms on broomrapes from this region should be indicators of potential pathogenic fungi. Furthermore, these pathogenic fungi should be able to infest different broomrape stages. The aim of my thesis is to contribute to the evaluation of an integrated control solution for this parasitic weed and focus on the use of pathogenic fungi isolated from symptomatic broomrapes as biocontrol solutions. A collection of 99 isolates was constituted. We studied the germination inhibition of broomrape seeds by each fungal isolate to investigate their impact on broomrape seeds in petri dishes. The pathogenicity of these fungi was assessed by inoculating them in pots containing soil, rapeseed and broomrape seeds to record the growth inhibition or necrosis induction of young broomrape stages. The 99 isolates showed a variable germination inhibition of broomrape seeds, from zero to 89% germination. One strain of Fusarium strongly inhibited broomrape germination and was significantly virulent when inoculated close to broomrape seeds leading to 84% of necrotized broomrapes in average compared to 1% necrotized broomrapes with another strain. However, this first strain was not as virulent when inoculated at the surface of the soil. The results highlight the importance of the inoculation mode, and the complexity of soil fungi, rapeseed and broomrapes interactions. Further cytological experiments would enlighten the colonization mode of the fungus to improve its pathogenic virulence against broomrape.