For over two decades, intraperitoneal administration of vancomycin and an aminoglycoside has been an accepted regimen for the empiric treatment of peritonitis in the peritoneal dialysis patient, until definite identification of the organism has been made. The recent emergence of vancomycin-resistant organisms has been of great concern in many centers. The current treatment recommendation therefore is to use cefazolin in place of vancomycin. We analyzed peritonitis data from January 1, 1996 to June 30, 1997, prior to switching over to cefazolin. Seventy-five percent (27 episodes) in 1997 as compared to 78% in 1996 were due to gram-positive organisms. Twenty-two percent (8 episodes) were due to gram- negative organisms in 1997, 21% in 1996, and 3% (1 episode) due to yeast in 1997, 3% in 1996. Staphylococcus epidermidis (SE) caused 33% of the gram-positive peritonitis episodes in 1997 as compared to 37% in 1996. Twenty-two percent of the gram-positive episodes were due to Staphylococcus aureus (SA) in 1997 and 46% in 1996. Enterococcal infections were 26% in 1997 and 1% in 1996. All of these were confined to only 1 patient. The antibiogram revealed 100% sensitivity of both SA and SE to vancomycin and 100% sensitivity of SA to cefazolin, but only 11% sensitivity of SE to cefazolin. The same patient population had a 48% sensitivity of SE to cefazolin in 1996, showing a sudden and substantial increase in resistance to SE. Even though SE is thought to be a less virulent organism, treating patients with a high probability of being infected by SE with an antibiotic showing 89% resistance is not warranted.