Research with older persons suffering from physical illness presents numerous challenges to gerontologists. Issues of conceptualization pertaining to the definition of illness, its location in the research paradigm, and the context in which illness occurs must be addressed prior to dealing with methodological problems. Access to physically ill elderly persons is often controlled by health care providers who serve as gatekeepers and/or referral sources in a broad range of studies. The suspicion and reluctance of gatekeepers may be overcome by establishing contacts prior to soliciting participation and by considering the needs and perspectives of the institution. Informed consent is often difficult to obtain from older adults due to deep-seated fears, restrictions imposed by families, and ethical considerations. Foremost among sampling issues is the difficulty of obtaining a randomized sample due to the potential threats of survivorship, health status, health utilization, and special population biases. Research process issues are hampered by the paucity of appropriate instruments and the effects of the patient's illness on both the process and the content of interviews. Data interpretation must take into account the generalizability of findings given the sampling difficulties with this population. To address these methodological research problems, a dynamic model, in which physical and mental health, the interview situation, contextual factors, and patient responses serve as input variables, is proposed. (BL)