The effects of date of inoculation on the development of Pierce's disease (PD) were evaluated in California grapevines during 1997 through 2000 at four locations. Some vines that had been inoculated either by using blue-green sharpshooters (Graphocephala atropunctata) as vectors or mechanically by needle puncture with the PD causal bacterium Xylella fastidiosa became infected during each month and at each location where infection was attempted. Vines inoculated on the earliest inoculation dates (April to May) developed more extensive and severe PD symptoms, and only 54% of these vines recovered from PD after the following winter, compared with vines that had been inoculated during June through August, of which 88% recovered from PD after the following winter. For the 1999 inoculations, the number of vines infected at a central California site (Parlier) was higher than the number of vines infected at a north coastal site (Hopland). For the best-fitting regression equation, percent recovery of vines infected with X. fastidiosa increased significantly with date of inoculation (r(2) = 0.737) at all sites excluding Hopland. The Hopland site had the highest percentage of vines that recovered from PD (100%). At most sites, only early infection (April and May) resulted in chronic disease unless the vines were inoculated at the bases instead of the distal tips of canes. Vines inoculated early in the growing season (April and May) have less chance to recover from Pierce's disease than vines inoculated later (July and August).