ObjectiveThis systematic review evaluates the use of cognitive load measurement methods in usability testing across diverse software interfaces. It provides guidance for researchers and practitioners by proposing a framework for selecting appropriate cognitive load measurement techniques.BackgroundCognitive Load Theory offers insights into software usability by addressing users' mental effort during task performance. Although cognitive load measurement methods are increasingly used in usability testing, no comprehensive analysis has focused specifically on various software interfaces.MethodWe systematically analysed 87 experimental studies published between 2001 and 2025. Databases searched included IEEE, ACM, ScienceDirect, SpringerLink, and Scopus. Inclusion criteria focused on studies applying cognitive load measurement to usability testing for different types of software interfaces.ResultsCognitive load measurement methods were categorised as subjective (e.g., NASA-TLX and self-reports) or objective (e.g., EEG, eye-tracking, dual-task paradigms, and physiological sensors). The most frequently used methods were performance measures (19%), NASA-TLX (12%), and eye fixations (11%). Commonly evaluated platforms included websites, virtual reality systems, and productivity tools. Each method's applicability, strengths, and limitations were identified.ConclusionThe review synthesises the relative merits of cognitive load measurement methods in usability evaluations and introduces a framework to guide the selection of techniques based on interface type and evaluation goals.ApplicationThe proposed framework operationalises CLT to support targeted, user-centred usability testing. It facilitates the selection of effective cognitive load measurement strategies, enhancing evaluation accuracy and informing better software design. Cognitive load theory is principally concerned with biologically secondary information. There is a specific cognitive architecture that allows humans to acquire, process, and use biologically secondary information ( Sweller, 2025 ). Within the framework of CLT, information acquisition refers specifically to how learners obtain biologically secondary knowledge that demands deliberate cognitive processing and the engagement of working memory. Within the CLT framework, this limitation is considered theoretical rather than absolute: when dealing with familiar or automated information stored in long-term memory, working memory is assumed to operate as if its capacity is no longer limited ( Paas, Renkl, & Sweller, 2004 ; Sweller, 2025 ). In such cases, information is retrieved as integrated “chunks,” reducing the load on working memory ( Sweller, 2010 ). This cognitive architecture provides the foundation for CLT and underpins how mental effort is managed during complex learning and problem-solving tasks. Learning is enhanced by minimising mental resources that are allocated to deal with a user interface and teaching materials so that working-memory can use its maximum resources to deal with the instructional topic ( Sweller, 2010 ). Therefore, when the nature of the software or the associated interface are cognitively demanding, a good user interface design or an appropriate and comprehensive help system become important ways to decrease extraneous cognitive load. Review Methodology We used the Kitchenham and Charters (2007) methodology, which is a widely adopted approach in computer science and human–computer interaction research. (2012) , in the context of using cognitive load measurement methods for interface and e-learning system design, subjective techniques assume that users can reliably monitor and report their cognitive processes while working with an interface. These methods are widely used to measure cognitive load as they are easy to conduct, they do not interfere with the primary task and they have proven to be reliable methods to indicate the amount of cognitive load ( Ayres, 2006 ). In order to increase the reliability of self-report methods, using multi-question assessment has been suggested ( Gerjets et al., 2009 ). For example, to evaluate the cognitive load of users while working with a website, we can use the tapping method by asking users to use one of their hands to control the mouse and the other hand to tap on the desk. If users face any usability issue such as an inability to find the desired feature, it is expected that the number of taps will be decreased as they should expend more mental effort to work with the website. After finding the desired feature, the number of taps should return to the normal baseline rate ( Miyake, Onishi, & Poppel, 2004 ). Linguistic analysis was used either independently or in conjunction with subjective rating questionnaires or performance measures (e.g., Khawaja, Chen & Marcus, 2012 , 2014 ; Oviatt, 2006 ; Tuček, Mount & Abbass, 2012 ). In this measurement, the interface element that is relevant for the current cognitive activity is identified based on the gaze direction ( Rudmann, McConkie, & Zheng, 2003 ). Increasing the fixation duration on an interface element can show that users needed more attention as the understanding of the interface was difficult for them. It can be an indication of increased processing in working memory and high cognitive load ( Chen et al., 2011 ). See Table 12 for relevant studies that used the fixations measurement method alone or in conjunction with other measurement methods. In addition to the above methods, since self-reported difficulty or mental effort questionnaires can provide valuable information and do not create any discomfort for the participants, they can be used as a complementary method at the completion of a usability test. Finally, in order to conduct an accurate usability test, a combination of subjective and objective cognitive load measurement methods should be utilised to compensate for the shortcomings of each method and increase the reliability of the results.