A library of arginine-like surface modifiers was tested to improve the targetability of DOPE:DOPC liposomes towards myofibroblasts in a tumor microenvironment. Liposomes were characterized using zeta potential and dynamic light scattering. Cell viability remained unchanged for all liposomes. Liposomes were encapsulated using doxorubicin with an encapsulation efficiency >94%. The toxicity of doxorubicin-loaded liposomes was calculated via half-maximal inhibitory concentration (IC50) for fibroblasts and myofibroblasts. These liposomes resulted in significantly lower IC50-values for myofibroblasts compared to fibroblasts, making them more toxic towards the myofibroblasts. Furthermore, a significant increase in cell internalization was observed for myofibroblasts compared to fibroblasts, using fluorescein-loaded liposomes. Most importantly, a novel regression model was constructed to predict the IC50-values for different modifications using their physicochemical properties. Fourteen modifications (A-N) were used to train and validate this model; subsequently, this regression model predicted IC50-values for three new modifications (O, P, and Q) for both fibroblasts and myofibroblasts. Predicted and measured IC50-values showed no significant difference for fibroblasts. For myofibroblasts, modification O showed no significant difference. This study demonstrates that the tested surface modifications can improve targeting to myofibroblasts in the presence of fibroblasts and hence are suitable drug delivery vehicles for myofibroblasts in a tumor microenvironment.