The aim of the present study was to evaluate the results of a prospective randomised trial comparing cephalic vein cut-down (CCD) versus percutaneous subclavian vein puncture with the Seldinger technique (PSP) in the implantation of subcutaneous permanent central venous access devices (SPCVAD) in order to better define the safer technique in outpatient surgery. Fifty patients were randomly divided into two groups: CCD--17 males, 8 females (age range: 35-75 years; mean: 60 years); PSP--17 males, 8 females (age range: 17-75 years; mean: 63 years). The two groups were compared in terms of implant morbidity, technical failure, operative time, patient acceptance, and healthcare costs. Data were analysed statistically using the chi-square test (P < 0.05 was considered significant) and Student's t-test. Technical failure occurred in 2 PSP patients (8%) and in 4 CCD patients (16%) (P = 0.663; chi 2 = 0.189). Mean operative time was 40 min in PSP (range: 35-70 min) and 50 in CCD (range: 35-60 min) (p = 0.108, T = -1.64). Patients were requested to define the pain experiences as a result of the operation as mild, moderate or severe: 23 PSP patients reported mild to moderate pain (92% of cases), and 2 patients severe pain (8%). The same figures were obtained in the CCD group, i.e. 23 (92%) and 2 (8%) patients, respectively (P = 0.377; chi 2 = 0.781). In our institute both types of implant cost $1260; in the case of PSP there is an additional cost of $120 for postoperative chest x-rays, making a total cost of $1380 per implant. The data of this randomised trial show that the implantation of a totally implantable vascular access device is a surgical procedure with a limited rate of morbidity in both cephalic vein cut-down and subclavian vein percutaneous puncture, with no statistically significant difference between the two in terms of associated morbidity, technical failure, operative time and patient acceptance.