ABSTRACT: A personal health record is an eHealth technology in which users can observe their progress over time for a given condition. A research gap was identified in the literature concerning the study of the amount of energy that these systems need for their operation, and the energy efficiency that may be attained depending on their design. After the selection of five representative personal health records, a total of 20 tasks commonly done, and based on previous work, were performed with regard to two proposed scenarios, namely patient use and health personnel usage. The power consumption of the main components of a host machine was measured during the performance of the proposed duties. To that end, a hardware tool called the Energy Efficiency Tester was employed. The data collected were analyzed statistically, and significant differences were found in the respective consumption of the display (χ2 (4) = 23.782, p = 0.000), the processor (χ2 (4) = 29.018, p = 0.000) and the whole PC (χ2 (4) = 28.582, p = 0.000). For all of these components, NoMoreClipBoard was the personal health record that required the least energy (57.699 W for the display, 3.162 W for the processor and 181.113 W for the whole PC). A total of two strong correlations were found in the energy consumption between the hard disk and the graphics card (r = 0.791, p < 0.001), and the processor and the PC (r = 0.950, p < 0.001). Some features generated special amounts of power consumption, such as the news wall found on PatientsLikeMe, or the use of load icons that had an impact on most PC components. In addition, an in-depth analysis of the user interfaces was performed. A discussion was carried out on the design of the user interfaces, also taking into account recommendations drawn from the literature, checking for their implementation in the personal health records selected. With the aim of promoting sustainability among software developers, a best practice guideline on sustainable software design was proposed. Basic sustainability recommendations were collected for professionals to consider when developing a software system in general, and a personal health record in particular.