Background: A clinical swallow examination (CSE) provides integral information that informs the diagnostic decision-making process within dysphagia management. However, multiple studies have highlighted a high degree of reported variability within the CSE process. It has been hypothesized that such variability may be the result of the clinical reasoning process rather than poor practices. Aims: To elucidate the nature of expert, speech-language therapists' (SLTs) clinical reasoning during an initial bedside assessment of patients referred for suspected dysphagia in the acute care environment. Methods & Procedures: An exploratory 'observation of practice' qualitative methodology was used to achieve the aim. Four expert SLTs, from two clinical services, completed CSEs with 10 new referrals for suspected dysphagia. All assessments were video-recorded, and within 30 min of completing the CSE, a video-stimulated 'think aloud' semi-structured interview was conducted in which the SLT was prompted to articulate their clinical reasoning at each stage of the CSE. Three types of concept maps were generated based on this video and interview content: a descriptive concept map, a reasoning map and a hypothesis map. Patterns that consistently characterized the assessment process were identified, including the overall structure; types of reasoning (inductive versus deductive), facts (i.e., clinical information) drawn upon; and outcomes of the process (diagnosis and recommendations). Interview content was examined to identify types of expert reasoning strategies using during the CSE. Outcomes & Results: SLTs' approach to clinical assessment followed a consistent structure, with data gathered pre-bedside, during the patient interview and direct assessment before a management recommendation was made. Within this structure, SLTs engaged in an iterative approach with inductive hypothesis-generating and deductive hypothesis-testing, with each decision-making pathway individually tailored and informed by patient-specific facts collected during the assessment. Clinical assessment was primarily geared towards management of an initial acute presentation with less focus on formulating a diagnostic statement. Conclusions & Implications: Variability in reported dysphagia practice is likely the result of a patient-centred assessment process characterized by iterative cycles of fact-gathering in order to generate and test clinical hypotheses. This has implications for the development of novel assessment tools, as well as professional development and education of novice SLTs.