We propose new systems for certificate revocation that are more economical and efficient than traditional ones. We also point out what we believe to be a structural problem in traditional public-key infrastructures, and various ways to solve it.