To evaluate the safety, feasibility, and clinical efficacy of percutaneous transhepatic injection of beta-cell grafts in patients with type 1 diabetes mellitus. Between December 2001 and November 2003, 15 patients with C-peptide-negative type 1 diabetes underwent 31 percutaneous injections for intraportal implantation of beta-cell grafts. Grafts consisted of cultured beta-cell preparations as previously described. In 13 cases, the transplant procedure was done under sedation, whereas in 18 cases, general anesthesia was given. In all procedures, percutaneous access to the right portal vein occurred under ultrasound (US) guidance with use of a microbore puncture needle. The subsequent catheterization of the main portal vein was performed under fluoroscopic and angiographic control with use of a microbore delivery catheter and guide wire. Clinical, biochemical, and radiologic evaluation was performed before and after the procedure. In all cases, it was possible to access the portal vein (median number of needle passes, 1; range, 1-6). The volume of cultured beta-cell grafts injected for each transplantation averaged 0.58 mL (range, 0.26-1.60 mL) and the mean recorded procedure time (from puncture to catheter withdrawal) was 19 minutes (range, 10-80 min). Three patients presented with transient abdominal pain immediately after the procedure; postprocedural duplex US of the liver revealed a patent portal vein and end branches in all cases and a minor perihepatic fluid collection in another three patients. From the end of week 1 to week 3, a mean 3.8-fold increase in liver aminotransferase levels was measured in all recipients after the first implantation session. A similar increase was seen in only one patient after a second transplantation session. At 6 months after transplantation, 13 of 15 patients (86%) had a functioning graft with plasma C-peptide levels greater than 0.5 ng/mL. The combined US, fluoroscopic, and angiographic monitoring of percutaneous transhepatic injection with use of a microbore delivery catheter is a safe and reproducible radiologic procedure for transplantation of beta-cell grafts in diabetic patients. Increased posttransplantation C-peptide levels, which demonstrate acceptable graft function, can be obtained.