The understanding and treatment of pain is one of the oldest challenges for physicians, scientists and philosophers. Much of our present rationale of pain control is based on the Cartesian idea that pain mostly originates from external or internal noxious stimuli, which are transmitted to and interpreted in the brain. Consequently, removal (blocking) of the stimuli and modification of cerebral awareness have been the prime targets of analgesic interventions. Only recently has the relationship between pain and other physical, psychological and social aspects of illness been considered in the overall management plan. Most of the literature on pain control reveals the physical bias of measurement. Apart from simple but reliable tools such as visual analogue scales and Likert-type verbal scales, more sophisticated measures such as multidimensional pain inventories have also been used when it is necessary to characterise pain more specifically. In clinical studies, it is usual to ask the patient to report on his own pain, although proxy measures such as mobility, performance status and analgesic consumption are also often used. The hospice concept of "total pain", in which the psychological, social, spiritual and other aspects are emphasised, has been influential in our new approach to pain measurement. Particularly when it is chronic and related to advancing disease as in metastatic cancer, pain can interact significantly with many facets of daily living. A holistic model of quality of life in such patients should, therefore, include a multidimensional or modular assessment of these areas. Medical interventions themselves can affect quality of life in both positive and negative ways. Some side-effects may be so common as to be accepted as "normal", e.g. constipation or sedation with opioids: it is only by their careful evaluation, when comparing opioids with essentially similar analgesic potentials, that differential toxicities may be revealed. Simple recording of physical side-effects of drugs is really not sufficient, because analgesics as well as other therapies may be associated with mood changes and broader consequences for quality of life. Only in the past few years has quality of life been seriously addressed in palliation research. For example, standardised quality of life scales are now included almost routinely in oncological studies involving radiotherapy or chemotherapy by the Medical Research Council (MRC) of Great Britain. Trials of the new biphosphonates, which can reduce bone pain in metastatic cancer, have been enhanced by incorporating quality of life measures. Based on the experience from earlier efficacy/safety studies with the new transdermal route of drug delivery for the opioid fentanyl, important areas of life such as sleep and cognitive function have been addressed. Randomised controlled trials of analgesics which include quality of life endpoints are still rare, but should be encouraged as these represent the most rigorous way of evaluating new therapies. The current preoccupation with quality assurance in healthcare is directed, ultimately, to the delivery of a better quality of care, which should also be more cost-effective, for large populations. An important intermediate step towards that ideal is the collection of data on pain and other symptoms, but also validated quality of life parameters on well-defined groups. Only by widening the scope of analgesic studies to include these dimensions can we hope to define appropriate strategies for more rational healthcare.