The key feature defining transcatheter cardiovascular interventions is that access to the vessels and heart is achieved by arterial puncture with a needle, rather than surgical incision with a scalpel. However, arteriotomy and vessel closure are performed without direct visualization of the arterial wall, which risks vessel damage and bleeding. Vascular closure devices offer the potential for enhanced control of access-site haemostasis and reduced complications in comparison with manual compression. However, although randomized clinical trials have shown reductions in time to haemostasis and ambulation, the data do not demonstrate consistent reductions in access-site complications or improvements in clinical outcomes. Another approach to increase the safety of percutaneous procedures is to use radial, rather than femoral, arterial access, a strategy that has polarized opinions among cardiologists. Clinical trial data show a clear reduction in access-site bleeding and complications with radial access, at the expense of a marginal increase in markers of procedural efficiency. However, randomized trials have not demonstrated improved clinical outcomes with radial access. The lack of impact on prognostically relevant bleeding events could explain this null finding, although the setting of primary percutaneous coronary intervention could be an exception. Ongoing, iterative improvement in catheter technologies, as well as in adjuvant antiplatelet and antithrombotic therapies, are likely to underlie the difficulty in demonstrating clear outcome benefits with different vascular access and closure strategies.