In order to improve the isolation of anaerobic bacteria from clinical materials, we have been using Stuart transport medium to hold swab specimens and phenylethanol blood agar to inoculate specimens since 1978. Also, all purulent materials have been cultured anaerobically regardless of the request. To evaluate the effect of such a practice, an analysis was made on the data obtained during the period of September 1978 to August 1979. The following results were obtained. 1) A total of 256 positive cultures of anaerobic bacteria were isolated from 187 patients. Of these 87 were gram-positive cocci and 143 gram-negative bacilli. 2) Anaerobes were most frequently isolated from specimens of abdominal, head and neck, and genital lesions. 3) The percentage of the number of specimens with both anaerobes and aerobes from the same specimen against the total isolation of anaerobes has increased from 41.4% in 1975-77 to 77.5% in 1978-79. 4) The percentage of anaerobe isolations against the total number of bacteriological specimens was 0.48% in 1975-77 and 1.34% in 1978-79. 5) The percentage of Clostridium against the total number of anaerobe isolates was 18% in 1972-73 and 3.1% in 1978-79. These facts clearly indicated that our new measures taken in 1978-79 have resulted in favorable effects in the isolation of non-sporeforming anaerobic bacteria from clinical specimens.