The unscheduled DNA synthesis (UDS) induced by ultraviolet radiations (UVR) in fibroblasts from 5 patients with Bloom Syndrome (BS) has been studied. Since often a high proportion of BS cells has large, probably polyploid nuclei, care was taken to select cells of normal size. Furthermore, UDS was usually measured over constant nuclear areas by a photometric method and each cell strain was tested on several different occasions. This showed that each BS cell strain had levels of UDS exceeding control values, on average, by 19-29%. The difference between the UDS of BS and control cells did not change with UVR doses from 5 to 100 J . m-2 but, unirradiated BS cells showed no evidence of "spontaneous" UDS. Also erythemal UVR (greater than 295 nm) elicited excessive UDS in the BS cell strain which was exposed to such radiations. When BS and control fibroblasts were incubated at low (32.5 degrees C) and high temperature (40.5 degrees C) before and after UVR doses of 5, 25 and 100 J . m-2, it was found that at the lowest dose the temperature of incubation did not modify the difference in UDS between normal and BS cells while, at the highest dose, BS cells incubated at 32.5 degrees C did not show more UDS than the controls. Finally, BS and control fibroblasts were fused with xeroderma pigmentosum (XP) cells of complementation groups C and D, which are complemented slowly by their partners, and it was than found that BS cells may transfer their tendency to high levels of UVR-induced UDS to their fusion partners. In keeping with these findings, UV-irradiated fibroblasts from BS heterozygotes seemed to show more UDS than normal cells. The anomalous behaviour of BS cells exposed to UVR is unexplained but trivial factors such as differences in cell geometry do not seem to account for our findings. Therefore, since an abnormally small endogenous pool of thymidine and "spontaneous" UDS in BS were not observed, it seems possible that the greater UDS performed during the first 1.5 h of repair by BS cells may be due to: greater numbers of damaged sites repaired; greater incorporation of thymidine per site repaired; or finally, both. In fact, the cell-fusion experiments suggest that BS cells may contain a diffusible factor which influences at least the initial rate of UDS. If such a factor were the product of the BS allele it could be argued that the BS mutations are not amorph and that the BS gene product may compete with that of the normal allele and modify the initial rate of UDS induced by UVR. It is hoped that our observations and their may possible interpretations will stimulate further work on BS.