The diverse topics that make up statistical computing present a challenge for constructing the appropriate courses in statistics curricula. An undergraduate course covering programming and data management prepares students for most statistics courses. However, many graduate students successfully learn these skills on their own; moreover, squeezing such a skills course into the master's program is difficult. We teach most of the topical statistical software in the courses that use them. For doctoral students, the issue becomes one of requirements and electives: some computing topics should be covered in the required courses, others are left to an elective. The selection and placement of these topics depend of the research program of the department. Even for an elective course, the research needs drive the choice of topics, although experience has shown that there can be no substitute for a good foundation in arithmetic and numerical linear algebra.