We've done B16 xenografts commercially (http://altogenlabs.com/xenograft-models/melanoma-xenograft/b16-xenograft-model/) and they are generally good for evaluating drug effectiveness based on growth curves. For metastatic models we stably-transfect the cells with GFP and track where they go throughout the mouse's body.
The B16 cell line that spontaneously forms tumor after chemical induction of melanoma in C57BL/6J mice and gives rise to a diverse spectrum of subclones with various propensities for proliferation, invasion, and metastasis. Two well-established subclones of the B16 melanoma cell line, obtained from in vivo passaging, include the B16F1 and B16F10 variants. B16F1 is characterized by low metastatic potential and, thus, is useful for studying primary tumor growth.69 Conversely, B16F10 has high metastatic potential to distant visceral organs, most notably the lungs, and has been ideal for in vivo studies because of its swift growth pattern and high turnover.