Eighteen patients with uncomplicated degenerative joint disease requiring joint replacement (hip or knee) were given a single 2-g intravenous dose of aztreonam over a 5-min period preoperatively. The mean concentration in synovial fluid of 83.0 9.2 >g/ml averaged 0.99 times the concomitant levels in serum. The mean concentration in cancellous bone of 16.0 4.3 Lg/g averaged 0.20 times the concomitant levels in serum.The increasing incidence of gram-negative bone and joint infections (14,19), particularly in patients with underlying diseases (13), combined with a higher incidence of poor therapeutic outcome in patients with such infections treated with available antimicrobial agents (10,29) (17), except that the samples were pulverized with a Spex freezer-mill prior to assay. The limit of detection of the assay was 0.1 ,ug/g of bone. The coefficient of variation and mean recovery from standards were 8.0 and 79.0%, respectively.Bone samples were prepared for high-pressure liquid chromatography by a procedure identical to that used for the bioassay and then filtered (Millex; 0.45 pum [pore size]; Millipore Corp., Bedford, Mass.). They were then analyzed with a high-pressure liquid chromatography system consisting of an Altex model 11OA pump set for 2 ml/min, a 37-to 50-,um silica gel precolumn, a 30-to 38-,um C18 guard column, a Waters Associates, Inc., 10-,um ,Bondpak analytical column (150 by 3.9 mm), and a UV detector set for 290 nm. The mobile phase was 85 parts 5 mM tetrabutylammonium sulfate, adjusted to pH 2.85 with 1 M KH2PO4, and 15 parts acetonitrile. Samples were introduced via a 20-,u precision loop injector. The limit of detection was 1 ,ug/g, the coefficient of variation was 8.1%, and the mean recovery from standards was 87.0%.The method used for microbiological assay of synovial fluid was identical to that previously reported for urine (20), except that samples were diluted 1:20 in 0.1 M phosphate buffer (pH 6) prior to assay. The limit of detection was 0.4 ,g/ml. The coefficient of variation and mean recovery were 3.5 and 98%, respectively. The high-pressure liquid chromatographic assay of synovial fluid was done by a method identical to that previously reported for serum (20). The limit of detection was 0.5 ,ug/ml. The coefficient of variation and recovery from standards were 8.5 and 98%, respectively.Assays of hemoglobin in blood, bone, and synovial fluid and correction of aztreonam concentrations for blood contamination were performed by previously described methods (17). Samples whose blood contents were 51% or more were excluded from analysis. No specimens suitable for assay could be obtained from patient 6.Results of the aztreonam assays are listed in Table 1. The mean concentration in serum at the time of bone and synovial fluid removal was 78 ,ug/ml. The mean concentration in bone was 16 ,ug/g. Concentrations of aztreonam in individual bone specimens were variable but showed a 710