Thirty-nine single antral biopsies (phase 1) and 99 sets of six antral biopsies (phase 2) were collected from 132 patients, and 87 (63%) yielded positive cultures for Campylobacter pylori. Of several primary media tested in phase 1, tryptic soy agar and Skirrow agar, each supplemented with 10% whole sheep blood, supported relatively good growth of C. pylori. In phase 2, four of the six biopsies in each set were tested with different urease systems. Selective urea agar for rapid identification was the most sensitive (39 of 63 [62%] at 1 h) and specific (100%); however, the difference between this system and the CLOtest was not statistically significant. The remaining two biopsies, one transported in saline and the other transported in a supplemented tryptic soy broth, were ground separately and inoculated onto tryptic soy agar and Skirrow agar, each supplemented with 10% whole sheep blood. In selected instances, 10% horse serum or 10% horse serum and 5 mM urea or 1% cholesterol were also added to the media. Smears stained with a modified Gram stain or acridine orange detected 68% of 63 culture-positive biopsies; no false-positive results were reported. Skirrow agar supported better growth of C. pylori than did tryptic soy agar; the nonselective medium was also overgrown with contaminants in 25 to 30% of the positive cultures. Based on colony size, Skirrow agar supplemented with 10% whole sheep blood, 10% horse serum, and 1% cholesterol supported optimal growth of C. pylori. Fresh media supported better growth than did prepoured commercial media (P less than or equal to 0.004). Saline was a convenient and satisfactory transport medium for antral biopsies.