BACKGROUND: Continuous monitoring blood culture systems (CMBCS) reduce the time and false negative rates of bacterial growth compared with the traditional manual blood culture systems which have been used in many hospitals yet. The purpose of this study is to evaluate the terminal subcultures monitored by Vital system compared with the manual system and to determine the guideline of terminal subcultures. METHODS: A retrospective study was conducted over a period of one year (from January to December 1995) with manual blood culture system and and sixteen months (from February 1996 to May 1997) with Vital system. All of the positive and negative blood bottles were done Gram staining and subcultured aerobically and anaerobically with 7-day terminal subculture protocol. All of the isolates were identified with API systems or ATB systems. RESULTS: Among 3,344 cases with the manual system, 305 cases (9.1%) were declared positive and 424 cases (8.8%) out of 4,822 cases with Vital system were positve. The terminal subcultures detected 48 cases (1.44%) in manual system and 9 cases (0.19%) in Vital system according to 7-day protocol. No statistical differences were observed in results among 5 day, 7 day and terminal subcultures. Those of false negative organisms were gram positive cocci (22 cases), Enterobacteriaceae (13 cases), non-fermenters (12 cases) and gram positive rod (1 case) with the manual system and gram positive cocci (4 cases), Entrobacteriaceae (1 case), non-fermenter (1 case) and yeasts (3 cases) with Vital system. CONCLUSIONS: These results suggest that terminal subculture of Vital systemnegative blood culture bottles is not necessary except S. aureus and fungus bacteremia on the basis of clinical situation.