Thymidylate synthase (TS) [TYMS; OMIM reference number (188,350)] is normally considered to be a cytoplasmic enzyme. However, a few reports have suggested it may also be present in the nucleus. To explore this in more detail, we used a highly specific polyclonal antibody to TS and a combination of techniques, including immunocytochemistry, confocal microscopy, cell fractionation, and Western blotting. We developed cell line HeLa-55, a HeLa derivative that grossly overexpresses TS. Although the vast majority of TS was in the cytoplasm, some TS also was seen in the nucleus. TS in parental HeLa cells and in normal human fibroblasts was seen exclusively in the cytoplasm. HeLa-55 cells exposed to 5-fluorodeoxyuridine were fractionated and examined by Western blotting. Interestingly, both free TS and the ternary complex of TS were seen in the cytoplasmic fraction but only free TS was detected in the nuclear fraction. Amongst different cell lines examined, HCT-15 and normal fibroblasts showed no nuclear TS, HCC-2998 and SW-620 showed a small amount of nuclear TS, and HT-29, RKO, and HCT-116 showed a strong nuclear TS signal. Nuclear staining was clearly evident in some clinical colorectal specimens, both normal and malignant. This staining was definitively shown to be TS by competition with recombinant TS protein. A putative leucine-rich nuclear export sequence was identified but its function could not be confirmed. We conclude that small amounts of TS protein is present in the nucleus of some cell types but further work is needed to determine the significance of this observation.