In a series of growth cabinet, glasshouse and field experiments, tissue samples from living clonal lines of Eucalyptus marginata (jarrah) were incubated immediately after sampling on agar (NARPH) selective for Phytophthora. Phytophthora cinnamomi was recovered 3-6 months after inoculation from 50% of samples with lesions and 30% of symptomless samples. However, up to 11% of samples with and without lesions and from which P. cinnamomi was not initially isolated contained viable pathogen. This was shown by removing tissue which had not produced any growth of P. cinnamomi on NARPH plates, cutting it into smaller sections, washing in sterile deionised water repeatedly for 9 days, and replating. Plating stem or bark tissue directly onto NARPH produced false-negative results for nine P. cinnamomi isolates and six jarrah clonal lines. The behaviour of the pathogen indicates that it could be present as dormant structures, such as chlamydospores, that need to be induced to germinate. Alternatively, fungistatic compounds in the tissue needed to be removed to allow the pathogen to grow. These results have important implications for disease diagnosis and management, disease-free certification and quarantine clearance.