Direct Raman imaging techniques are demonstrated to study the drug distribution in living cells. The advantage of Raman imaging is that no external markers are required, which simplifies the sample preparation and minimally disturbs the drug mechanism during imaging. The major challenge in Raman imaging is the weak Raman signal. In this study, we present a Raman image model to describe the degradation of Raman signals by imaging processes. Using this model, we demonstrate special-purpose image-processing algorithms to restore the Raman images. The processing techniques are then applied to visualize the anticancer agent paclitaxel in living MDA-435 breast cancer cells. Raman images were obtained from a cancer cell before, during, and after drug treatment. The paclitaxel distribution illustrated in these images is explained by means of the binding characteristics of the paclitaxel and its molecular target-the microtubules. This result demonstrates that direct Raman imaging is a promising tool to study the distribution of a drug in living cells.