INTRODUCTION: The functionality of subcutaneous venous access devices is evaluated at least every time the device is accessed. This evaluation is subjective and actions are only undertaken when blood withdrawal and/or fluid injection become problematic. The function of these devices has rarely been evaluated objectively. The present study tried to find an objective and standardized way to evaluate the withdrawal speed of a newly inserted port. METHODS: Between October 2001 and December 2002, a prospective randomized study of 3 types of ports was carried out. The ability to infuse heparinized normal saline and to withdraw blood was evaluated by recording the filling time of a 10 ml Vacutainer tube in 876 newly inserted central venous ports at the end of the surgical procedure. RESULTS: The patient groups were comparable in age, gender and insertion procedure characteristics (vein used, position of the port on the body and length of the catheter). The median time needed to fill a 10 ml Vacutainer tube in all ports was 17.00 sec (range 11.43-63.62 sec). The median filling time for BardPort was 16.36 sec (range 13.48-39.00 sec), for Celsite 18.35 sec (range 12.03-40.00 sec) and for Port-a-cath 16.43 sec (range 11.43-63-62 sec). A significant difference in filling time was found between the large bore catheters (BardPort, Port-a-cath) and the small bore catheter (Celsite) of 2 seconds median value. CONCLUSION: Measurement of withdrawal speed provides an objective criterion for the quality description of blood withdrawal immediately after insertion. A value of more than 20 seconds for filling a 10 ml Vacutainer tube could therefore be suggested a useful trigger for further investigation.