In this study, the addition of monoolein to phosphatidylcholine (PC), tricaprylin, and propylene glycol (PG) mixtures was studied to produce fluid precursor formulations (FIPs) that could transform into hexagonal phase (resistant to aqueous dilution) in vitro and in vivo. The overall goal was to obtain FIPs that could incorporate chemopreventive drugs for subcutaneous administration in the mammary tissue to inhibit the development and/or recurrence of breast cancer. Increasing PG content reduced FIP viscosity up to ~ 2.5-fold, while increases in PC (over monoolein) increased the formation of emulsified systems. The hexagonal phase was observed at 20% of water and higher, with the minimum amount of water necessary for this formation increasing with PG content. The selected FIP formed a depot in vivo after ~ 24 h of administration; its structure was compatible with the hexagonal phase and it remained in the mammary tissue for at least 30 days, prolonging the permanence of a fluorescent probe. In vitro, the release of the synthetic retinoid fenretinide was slow, with ~ 9% of the drug released in 72 h. Consistent with this slow release, fenretinide IC in breast cancer cells was ~ 100-fold higher in the selected FIP compared to its solution. The FIP reduced cell migration and presented higher cytotoxicity towards tumor compared to non-tumor cells. Given the limited number of options for pharmacological prevention of breast cancer development and recurrences, this formulation could potentially find applicability to reduce the frequency of administration and improve local concentrations of chemopreventive drugs.