Consumer eHealth interventions are of growing importance in individual management of health and health behaviors. However, a range of access, resource, and skill-related barriers prevent healthcare consumers from fully engaging in and benefiting from the spectrum of eHealth interventions. eHealth skills, such as familiarity with computers and grasp of health concepts, can facilitate effective use of eHealth tools. Cognitive skills range in complexity, and different eHealth tools will exert different demands on skill and cognition. We propose a framework and method for characterizing complexity of eHealth tasks, which can be used to diagnose and describe barriers encountered as well as inform the development of solution strategies. The framework integrates eHealth literacy with cognitive processes, providing a systematic method of characterizing demands of eHealth tasks. The method implements a novel approach to cognitive task analysis combined with user observations. The method aims to identify and characterize barriers encountered during user performance of the task. We applied the framework and method to the analysis of 6 web-based consumer eHealth tasks. The tasks are information seeking tasks, with some decision points, across a range of different health domains. We conducted observations and in-depth analysis of the task performance of 20 users on the same set of tasks. The task analysis was able to characterize task demands, providing a systematic way to characterize and measure the magnitude of different literacy and complexity demands. Analysis of task performance revealed the steps that were problematic for most participants, and the knowledge and skill gaps that led to barriers. The aggregate results identified information literacy to be a prominent barrier to task completion for most participants; these barriers describe struggles with recognizing and locating information cues and with evaluating the relevance of resources. We conducted further analyses to address emergent questions. Cluster analysis was applied to discern the underlying patterns of task demands and participant backgrounds across groups of questions and groups of participants. This analysis explored the factors that contribute to difficulty and variation in performance. The clusters revealed that literacy types play a greater role than cognitive processes in reducing task scores. The analysis also identified education level and having had experience searching for health information online to be common factors among clusters of participants with high task scores. We furthered analyzed task performance employing an in-depth analysis of protocols to characterize the information seeking process. The analysis revealed distinct differences in task performance between high and low scorers throughout the different stages of the information seeking process. The results support the value of the framework for characterizing task demands and for diagnosing and explaining barriers encountered in task completion. The novel framework and micro-analytic approach contributes new knowledge about eHealth competencies and can inform solution strategies to bridge the knowledge- and skill-related barriers.