Epidemics of phony peach disease (PPD), caused by Xylella fastidiosa, are of increasing concern to peach (Prunus persica) producers in the southeastern United States. Primers suitable for both conventional PCR (cPCR) and quantitative PCR (qPCR), along with optimal tissue and sampling time, are needed for comparative and reliable detection of X. fastidiosa. In this study, we developed and assessed novel primers for X. fastidiosa and for peach and compared detection of X. fastidiosa in four peach tissue types sampled at three time points using both cPCR and qPCR. Primer C06Xf-bamA was extensively tested for reliable detection of X. fastidiosa due to the more consistent intensity of the cPCR products and the marginally lower average quantification cycle (Cq) values of the qPCR products, compared with the other primers screened. Among the four peach tissue types tested, only root samples demonstrated reliable and consistent detection of X. fastidiosa; stem, petiole, and leaf samples, regardless of source trees, primers used, sampling times, or PCR methods (cPCR or qPCR), were unreliable for detection, due to insufficient quantity of DNA of X. fastidiosa in these samples based on the relative quantification assay. The Cq means and ratios were compared and statistically analyzed, to ascertain effects of source tree, tissue type, sampling time, and primer. Differences in detection sensitivity and the Cq means among sampled trees, sampling times, tested primers, and tissues (except root) were not significant or were inconsistent precluding further exploitation. In summary, these novel primers are a useful resource for detecting X. fastidiosa, and based on our results, root is the only tissue type reliable for year-round detection of X. fastidiosa in peach. Further research on potential utilization of above-ground tissues for PCR detection of X. fastidiosa are discussed.