The existence of a bystander effect following both alpha and gamma irradiation of many cell lines is not now in dispute. The significance of this effect for cancer risk assessment and radiotherapy treatment planning requires demonstration of its relevance in vivo. The problem in demonstrating the existence of the effect in vivo is that other systemic effects may mask or confound the effect being investigated and it is practically impossible to attribute an effect in a particular cell to a signal produced in another irradiated cell. To approach this problem, we have developed an assay where fragments of human tissue can be irradiated ex vivo and the media harvested and added to unirradiated, allogenic explants or to a clonogenic cell line which has a well characterized and stable response to the bystander signal. The variation in production of the signal from patient to patient can thus be assessed using molecular and cellular endpoints. A study using tissue from over 100 patients and from mouse strains with well characterized responses to low level radiation exposure shows that there is variation in the effect of the signal produced by irradiated tissue from different patients. Gender, smoking status and the existence of a bladder malignancy influence the expression of the signal by normal urothelium. The effects of exposure to medium containing the signal are transmitted to distant progeny of the exposed cell population. The results may be important not only for understanding radiation risk mechanisms for protection but also for radiotherapy treatment planning where they may open new avenues for development of drugs for combined therapy.