Xylella fastidiosa infects many different host plants and some present a challenge for its reliable detection and quantification in both, diagnostics and research. To elucidate the inhibition arising from different host plants and the efficiency of molecular detection methods, altogether 129 samples of 29 genera were spiked with a low concentration of X. fastidiosa cells. The recovery of X. fastidiosa was determined with digital PCR after automated magnetic beads DNA extraction, and the results compared with the results of real-time PCR. Overall, real-time PCR was less sensitive and less resistant to inhibition than dPCR, in particular in Rosmarinus, Lavandula, Origanum and Prunus samples that necessitate analysis of diluted DNA. For samples of olives (Olea europaea), critical samples extensively tested for X. fastidiosa in Europe, high inhibitory effect was not alleviated with analysis of dilutions and dPCR revealed low recovery of X. fastidiosa DNA in addition to inhibitors. In olives and other samples with high inhibition and/or low DNA recovery, dPCR provided a more reliable and accurate detection and quantification method than real-time PCR at lower number of technical repeats.