A group of 3'-O-butanoyl, 5'-O-butanoyl, and 3',5'-di-O-butanoyl esters of 5-fluoro-2'-deoxyuridine (FDU), and 2',5-difluoro-2'-deoxyuridine (DFDU), 3'-O-retinoyl, and 3',5'-di-O-retinoyl esters of FDU, and 5'-O-bis(2,2,2-trichloroethyl)phosphoryl-FDU and its 3'-O-butanoyl ester, was synthesized. These compounds were designed to act as double prodrugs that would serve as a depot to release two active drugs that act through different mechanisms. Thus, a nucleotide derivative of FDU or DFDU could act as a competitive inhibitor for thymidylate synthase, whereas retinoic acid and butyric acid would be expected to induce cell differentiation. The in vitro anticancer activities for these prodrugs were determined against a panel of nine tumor types (leukemia, non-small cell lung, colon, CNS, melanoma, ovarian, renal, prostate, breast) that encompassed about 60 human tumor cell lines. Structure-activity relationships indicate that O-butanoyl esters of FDU are approximately equipotent to FDU, the O-butanoyl esters of DFDU are less active than FDU, and the retinoyl and bis(2,2,2-trichloroethyl)phosphate derivatives of FDU exhibit comparable activity to FDU. In addition to their cytotoxic effect, 3'-O-retinoyl-FDU (12) and 3'-O-butanoyl-5'-O-bis(2,2,2-trichloroethyl)phosphoryl-FD U (16) also induced in vitro cell differentiation of promyelocytic leukemia HL60 cells. These combined cytotoxic and cell differentiation effects exhibited by 12 and 16 produced greater morphological drug-induced granulation and neutrophil vacuolation, and more cell apoptosis, than observed upon exposure to either retinoic acid or sodium butanoate. Dose-escalation studies in mice showed that 12 or 16 did not induce any acute or chronic toxicity, change in plasma clinical chemistry parameters, or gross histapathological changes at 60 days following an initial dosage regimen of 0.013 mmol/kg i.p. for 7-consecutive days. The in vivo growth delay response of murine mammary EMT6 solid tumors suggests that 3'-O-retinoyl-FDU (12) delays tumor growth relative to the other prodrugs investigated, sodium butyrate, retinoic acid, FDU, or a combination of retinoic acid and FDU. These preliminary results suggest that 3'-O-retinoyl-FDU (12) warrants further in vivo investigation to determine its tissue biodistribution and pharmacokinetic parameters that would be of value in assessing its potential usefulness as an anticancer prodrug.