Small-sized liposomes have several advantages as drug delivery systems, and the ethanol injection method is a suitable technique to obtain the spontaneous formation of liposomes having a small average radius. In this paper, we show that liposomal drug formulations can be prepared in situ, by simply injecting a drug-containing lipid(s) organic solution into an aqueous solution. Several parameters should be optimized in order to obtain a final suitable formulation, and this paper is devoted to such an investigation. Firstly, we study the liposome size distributions determined by dynamic light scattering (DLS), as function of the lipid concentration and composition, as well as the organic and aqueous phases content. This was carried out, firstly, by focusing on POPC (1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine) then on the novel L-carnitine derivative PUCE (palmitoyl-(R)-carnitine undecyl ester chloride), showing that it is possible to obtain monomodal size distributions of rather small vesicles. In particular, depending on the conditions, it was possible to achieve a population of liposomes with a mean size of 100 nm, when a 50 mM POPC ethanol solution was injected in pure water; in the case of 50 mM PUCE the mean size was around 30 nm, when injected in saline (0.9% NaCl). The novel anticancer drug Gimatecan, a camptothecin derivative, was used as an example of lipophilic drug loading by the injection method. Conditions could be found, under which the resultant liposome size distributions were not affected by the presence of Gimatecan, in the case of POPC as well as in the case of PUCE. To increase the overall camptothecin concentration in the final liposomal dispersion, the novel technique of "multiple injection method" was used, and up to a final 5 times larger amount of liposomal drug could be reached by maintaining approximately the same size distribution. Once prepared, the physical and chemical stability of the liposome formulations was satisfactory within 24, as judged by DLS analysis and HPLC quantitation of lipids and drug. The Gimatecan-containing liposomes formulations were also tested for in vitro and in vivo activity, against the human nonsmall cell lung carcinoma NCI-H460 and a murine Lewis lung carcinoma 3 LL cell lines. In the in vitro tests, we did not observe any improvement or reduction of the Gimatecan pharmacological effect by the liposomal delivery system. More interestingly, in the in vivo Lewis lung carcinoma model, the intravenously administration of liposomal Gimatecan formulation showed a mild but significant increase of Tumor Volume Inhibition with respect to the oral no-liposomal formulation (92% vs. 86 %, respectively; p < 0.05). Finally, our study showed that the liposomal formulation was able to realize a delivery system of a water-insoluble drug, providing a Gimatecan formulation for intravenous administration with a preserved antitumoral activity.