Aspects of the reservoir competence of four rodents for the Lyme disease spirochete, Borrelia burgdorferi, were evaluated in California. Rodents were live-trapped and ear-punch biopsies were cultured during each season. A second set of biopsies was cultured from representative individuals after 2–3 weeks of captivity and the results of culturing biopsies taken on both dates were compared with the results of feeding Ixodes pacificus larvae on hosts xenodiagnostically. The prevalence of infections did not differ significantly between dusky-footed woodrats (Neotoma fuscipes) and California kangaroo rats (Dipodomys californicus) nor among seasons. Combined results of the three tests showed that 85.7% of dusky-footed woodrats (n = 21) and 78.6% of California kangaroo rats (n = 14) were infected with B. burgdorferi. In contrast, only 22.2% of brush mice (Peromyscus boylii) (n = 14) and 7.1% of pinyon mice (P. truei) (n = 9) were infected. The sensitivity of culturing ear-punch biopsies as an assay for borrelial infection was significantly greater when biopsies were taken after a short period of captivity (0.89) rather than on the day of capture (0.52). Tick xenodiagnosis, in which I. pacificus was used as the vector, revealed borrelial infections in 90.3% of infected rodents. Spirochetes were observed in 37.7% of 239, 45.2% of 155, 60.0% of 10, and 7.1% of 14 cultures of nymphal I. pacificus fed as larvae on naturally infected woodrats, kangaroo rats, brush mice, and a pinyon mouse, respectively. The mean prevalence of infection in xenodiagnostic ticks varied significantly among host species with a greater proportion of ticks infected while feeding on woodrats and kangaroo rats than on mice. This study reconfirms previous reports that implicated woodrats and kangaroo rats as reservoirs of B. burgdorferi in California.