Translational sciences increasingly emphasize the measurement of functions in native human tissues. However, such studies must confront variations in patient age, gender, genetic background and disease. Here, these are discussed with reference to neuromuscular and neurosecretory functions of the human gastrointestinal (GI) tract. Tissues are obtained after informed consent, in collaboration with surgeons (surgical techniques help minimize variables) and pathologists. Given the difficulties of directly recording from human myenteric neurones (embedded between muscle layers), enteric motor nerve functions are studied by measuring muscle contractions/relaxations evoked by electrical stimulation of intrinsic nerves; responses are regionally dependent, often involving cholinergic and nitrergic phenotypes. Enteric sensory functions can be studied by evoking the peristaltic reflex, involving enteric sensory and motor nerves, but this has rarely been achieved. As submucosal neurones are more accessible (after removing the mucosa), direct neuronal recordings are possible. Neurosecretory functions are studied by measuring changes in short-circuit current across the mucosa. For all experiments, basic questions must be addressed. Because tissues are from patients, what are the controls and the influence of disease? How long does it take before function fully recovers? What is the impact of age- and gender-related differences? What is the optimal sample size? Addressing these and other questions minimizes variability and raises the scientific credibility of human tissue research. Such studies also reduce animal use. Further, the many differences between animal and human GI functions also means that human tissue research must question the ethical validity of using strains of animals with unproved translational significance.