Power and sample-size problems cannot be solved by a statistician in isolation; it is crucial that the investigator (usually a non-statistician) play the leading role, so that judgments about appropriate effect sizes and practical sample sizes are available. This article concerns graphical user interfaces that can facilitate the process. With the right add-in functions, a spreadsheet such as Microsoft Excel can be quite effective. In situations where several interrelated tests need to be examined at once, more sophisticated software is helpful. In the latter case, the Java object model can be used to great advantage.