This paper describes how we have used material science, physical chemistry, and some luck, to design a new thermal-sensitive liposome (the low temperature sensitive liposome (LTSL)) that responds at clinically attainable hyperthermic temperatures releasing its drug in a matter of seconds as it passes through the microvasculature of a warmed tumor. The LTSL is composed of a judicial combination of three component lipids, each with a specific function and each affecting specific material properties, including a sharp thermal transition and a rapid on-set of membrane permeability to small ions, drugs and small dextran polymers. Experimentally, the paper describes how bilayer-concentration changes involving the lysolipid and the presence or absence of DSPE–PEG2000 affect both the lipid transition temperature and the drug release. While the inclusion of 4 mol% DSPE–PEG2000 raises the transition temperature peak (T_m) by about 1 °C, the inclusion of 5.0, 9.7, 12.7 and 18.0 mol% MSPC slightly lowered this peak back to 41.7 °C, while not further broadening the peak breadth. As for drug release, in the absence of MSPC, the encapsulated doxorubicin-citrate is hardly released at all. Increasing the composition of MSPC in the lipid mixture (5.0, 7.4, 8.5 and 9.3 mol% MSPC) shows faster and faster initial doxorubicin release rates, with 8.5 and 9.3 mol% MSPC formulations giving 80% of encapsulated drug released in 4 and 3 min, respectively. The Thermodox® formulation (9.7 mol% MSPC) gives 60% released in the first 20 s. The presence of PEG–lipid is found to be essential in order for the lysolipid-induced permeability to reach these very fast times. From drug and dextran release experiments, and estimates of the molecular and pore size, the conclusions are that: in order to induce lasting nanopores in lipid bilayers 10 nm diameter, they initially require the presence (from the solid phase structure) of grain boundary defects at the DPPC transition and the permeabilizing component(s) can either be a pore forming lysolipid/surfactant plus a PEG–lipid, or can be generated by a PEG–surfactant incorporated at 4–5 mol%. The final discussion is centered around the postulated defect structures that result in membrane leakage and the permeability of doxorubicin and H⁺ ions.