The efficiency of three diagnostic methods, i.e. PCR, real-time PCR and LAMP, for detection of Xylella fastidiosa (Xf) genomic DNA from Philaenus spumarius (Ps) and Neophilaenus campestris (Nc) insect vectors was evaluated using three total nucleic acids (TNA) extraction methods (EM). In addition, a new real-time LAMP technology, Fluorescence of Loop Primer Upon Self Dequenching-LAMP (FLOS-LAMP), originally developed for human virus diagnoses, was optimized and assessed for detection of Xf in insect vectors. EM1 consisted of entire insects heated in an extraction buffer (EB) containing Tris-EDTA and TRITON-X100. In EM2, TNAs were extracted only from excised heads of insects, and heated again in the EB of EM1. EM3 consisted of grinding entire insects, heads and bodies recuperated from EM2, with a CTAB buffer. The molecular analyses conducted on 100 specimens of Ps and 50 of Nc, collected from a Xf-infected olive orchard (Lecce province, Italy), showed that 29% of specimens (40 Ps and four Nc) were positive to the presence of Xf. The comparison between the three methods revealed that EM3 is the most efficient for extracting Xf-genomic DNA from insect vectors, of which 44 specimens were positive for Xf in each of the diagnostic methods used, including the newly optimized FLOS-LAMP assay. In general, the real-time PCR and LAMP assays were more competent than the conventional PCR for detection of Xf in insect vectors, independently from the EM used. The newly optimized FLOS-LAMP technique had a detection limit of 1 fg mu L-1 of Xf-genomic DNA, compared to the 10 fg mu L-1 for conventional LAMP. The high sensitivity of the FLOS-LAMP was evident through the greater number of overall Xf-infected insect vectors detected (60%), compared to those for LAMP (45%,), real-time PCR (28%) and PCR (10%). FLOS-LAMP, being a more sensitive and specific assay, together with EM3, were the most appropriate approaches for an accurate detection of Xf in insect vectors.