The way in which physicians are trained to do invasive practical procedures is an ongoing challenge for educators. Percutaneous insertion of a central line via the femoral vein using the Seldinger technique is an important practical pediatric procedure, and the need for physicians to be educated in the necessary skills is recognized in current training initiatives such as Pediatric Advanced Life Support (PALS) and Advanced Pediatric Life Support. Unfortunately, the majority of instruction in central venous access techniques is theoretic. This approach does not provide the hands-on training needed to give practitioners the necessary practical experience, or confidence in their skills. Practice using simulated tissue can enable physicians to perform practical skills with greater confidence. However, although commercially available models exist for peripheral venous access, a recent cross-Canada survey of the 13 PALS program coordinators and a similar inquiry to the American Heart Association indicated that none of them had a pediatric practice model for central venous access. We describe 1) how to construct from materials readily available a pediatric model for the insertion of central venous catheters into the femoral vein using the Seldinger technique, and 2) an evaluation of the change in confidence learning with the model engendered. In our experience, this model is inexpensive (less than $50) and can be replicated readily by others for use as a teaching aid. It provides inexperienced physicians the opportunity to learn the practical elements of the technique and acquire confidence in the Seldinger method. Our hypothesis was that the confidence and skill of physicians would be increased by practical experience of central line insertion using a realistic model. The model enables trainees to be taught the technique described in the PALS manual to locate the femoral artery. They then can learn to introduce a thin-walled needle or over-the-needle catheter, one finger's breadth below the inguinal ligament and just medial to the location of the femoral artery. The needle or over-the-needle catheter then can be advanced at the correct angle if the needle is directed toward the model's umbilicus. As occurs in vivo, the model allows for a free flow of fluid to be obtained as the "vessel" is entered. If the Foley catheter simulating the vessel is transfixed, negative pressure applied as the needle is withdrawn will result in fluid being obtained as the needle tip reenters the "vessel." The syringe then can be removed from the needle, and the key elements of the procedure-correct insertion of the Seldinger guide wire and passage of the venous catheter over the guide wire into the vessel-can be practiced. If desired, instruction also can be given on the use of a dilator and techniques of taping the catheter in place and all the appropriate techniques to avoid potential air embolism. However, the model does not lend itself to instruction in suturing. The model has been used to teach the practical elements of this technique to 428 physicians (emergency physicians, 49%; pediatricians, 24%; other physicians, 20%; pediatric residents, 7%). Their success rate for cannula insertion in three or fewer attempts was 87%. The last 218 physicians were evaluated to assess the influence of learning with the model on their confidence to perform the technique successfully in an emergency. Before training they were asked, "Have you done a pediatric resuscitation course that taught this technique in theory?" and "Rate your confidence level for performing central vascular access in a patient from 0 to 5 (none, very little, some, moderate, good, complete)." This rating was repeated after the training session using the model. For 154 (71%) answering "yes" to a previous resuscitation course, mean scores were 1.52 (standard error [SE] +/- 0.91) after theoretic instruction and 4.06 (SE +/- 0.47) after practical education using our model. The 64 (29%) physicians