The detection and enumeration of circulating tumor cells (CTCs) has shown significant clinical utility with respect to prognosis in breast, colorectal and prostate cancers. Emerging studies show that CTCs can provide pharmacodynamic information to aid therapy decision making. CTCs as a 'virtual and real-time biopsy' have clear potential to facilitate exploration of tumor biology, and in particular, the process of metastasis. The challenge of profiling CTC molecular characteristics and generating CTC signatures using current technologies is that they enrich rather than purify CTCs from whole blood; we face the problem of looking for the proverbial 'needle in the haystack'. This review summarizes the current methods for CTC detection and enumeration, focuses on molecular characterization of CTCs, unveils some aspects of CTC heterogeneity, describes attempts to purify CTCs and scans the horizon for approaches leading to comprehensive dissection of CTC biology.