In the past decade there has been increasing attention to the public health importance of injuries. Public health agencies seeking to reduce injuries need methods for counting injuries, calculating injury rates, identifying the causes of injuries, and measuring outcomes of interventions. All of these areas present problems for injury epidemiologists and injury prevention programs. This paper provides a framework for classifying these problems into five categories: (a) numerator problems; (b) denominator problems; (c) causation; (d) exposure measurement problems; and (e) multiplicity. For most problems, we identify proven or potential solutions, using examples from the literature of injury epidemiology.