End-stage renal failure is a life-threatening condition, often treated with home-based peritoneal dialysis (PD). PD is a demanding regimen, and the patients who practise it must make numerous lifestyle changes and learn complicated biomedical techniques. In our experience, the renal nurses who provide most PD education frequently express concerns that patient compliance with their teaching is poor. These concerns are mirrored in the renal literature. It has been argued that the perceived failure of health professionals to improve compliance rates with PD regimens is because ‘compliance’ itself has never been adequately conceptualized or defined; thus, it is difficult to operationalize and quantify. This paper examines how a group of Australian renal nurses construct patient compliance with PD therapy. These empirical data illuminate how PD compliance operates in one practice setting; how it is characterized by multiple and often competing energies; and how ultimately it might be pointless to try to tame ‘compliance’ through rigid definitions and measurement, or to rigidly enforce it in PD patients. The energies involved are too fractious and might be better spent, as many of the more experienced nurses in this study argue, in augmenting the energies that do work well together to improve patient outcomes.