The practice of medicine is complex and challenging. Medical practitioners require a range of capabilities to practice effectively and consistently over a long career. Selection for the capacity to develop these capabilities and education to build them are becoming increasingly important in a rapidly-changing society that demands a return on its investment in medical schools and post-registration training. Although the range of capabilities required for trainee and graduate doctors is described in the medical student selection and education literature, the viewpoints of privileged stakeholders have generally been favoured in determining them, relative to those most affected by medical practice (and malpractice). This study attempts to address this question from a more inclusive stakeholder perspective, linking ‘capabilities’ to ‘doctor experiences’ in ways not previously reported in independently-refereed research literature. A qualitative phenomenological approach offered the most suitable paradigm for exploring the ‘doctor experience’. A pilot study was conducted to refine the research question and to inform the main study. For the main study, a total of 107 participants were purposively recruited to represent the diversity of stakeholders in medical practice. Thirtynine (39) one-on-one interviews were conducted with doctors (13), patients (11) and participants identified as ‘key stakeholders’ (15) because of their medical student selection and education expertise and experience. The remaining 68 stakeholders each participated in one of eight group interviews (as health professionals, medical educators, academics, health executive managers, medical students, patients or community representatives). The interviews focused on first-person accounts of stakeholders’ experience of their interactions with medical practitioners, which were audio recorded, transcribed verbatim, anonymised and imported into computer assisted qualitative data analysis software. The analytical methodology utilised was based on Smith’s (1996) Interpretative Phenomenological Analysis (IPA). In order to accommodate the larger-than-usual number of participants for an IPA study, Smith’s method was somewhat modified. All of the data were first analysed descriptively to identify broad categories of experience with doctors and the extent to which particular participants had provided vivid first-person accounts of those experiences. Participants whose narratives were the most deeply experiential and provided the richest accounts of particular doctor capabilities or concerning behaviours were designated ‘signature’ stakeholders for that category and their transcripts were interrogated with the full IPA methodology. This involved meaning-making and interpretation through application of the ‘hermeneutic cycle’. The analysis of these transcripts was then supported by less intensive interpretation of accounts from other stakeholders who had made sense of their experiences in similar ways. Broadly speaking, experiences of doctors were evaluated by participants as either positive or negative. Positively evaluated experiences focused on doctor capabilities in the major domains of being ‘patient-centred’ and being a ‘good communicator’. More specifically, these emphasised ‘ensuring patient safety’, ‘showing genuine concern for patients’, ‘combining competence with caring’, ‘being a good listener’, ‘allowing sufficient time’ and ‘being respectful’. Negatively evaluated doctor experiences focused on behaviours in the major domains of being ‘unprofessional’ and ‘compromising the safety of self and others’. More specifically, these related to ‘being disrespectful’, ‘being arrogant’, ‘having impaired judgement’, ‘being a poor communicator’ and ‘being incompetent’. The study’s findings were somewhat concordant with the existing literature. My participants’ positively and negatively evaluated experiences pointed to capabilities such as a patient-centredness and effective communication that had previously been identified as desirable in competency frameworks derived from the opinion of ‘privileged’ stakeholders such as ‘CanMEDS’ and the Australian Medical Council Graduate Outcome Statements. The study adds to the literature, however, in that the positively and negatively evaluated experiences of the broader range of ‘less-privileged’ stakeholders I interviewed prioritised different patterns of capabilities. They emphasised particularly patientcentredness, good communication, professionalism and practicing safely, characterising each of these domains in novel ways. Potential limitations of the study related to the novelty of the modification to Smith’s IPA method that I developed in order to balance respect for the idiographic (distinct and detailed) accounts of my participants with the need to manage data from a larger-than-usual number of participants for a phenomenological study. The modified method appeared to serve its purpose, however, and the range of measures taken to ensure the trustworthiness of its findings is described. My experience as a health professional and medical educator may have affected my degree of ‘reflexivity’ or ability to self-examine the potential implications of these experiences for my research role. It may also have influenced how participants viewed me, particularly those who were known to me previously. A further range of measures taken to overcome this potential limitation is also described. Implications of the study are discussed and recommendations made for future medical selection and education research and practice. These include strategies to refine the range and prioritisation of practitioner capabilities, on the basis of the experience and needs of a broader range of stakeholders, to enable doctors to practice effectively in today’s society. The study adds to the existing body of medical selection and education research. It also gives stakeholders an opportunity to be heard, particularly those most affected by medical practice and malpractice.