The present invention provides a method for determining a response of a mammalian subject having melanoma tumor cells to treatment with a melanoma inhibitory agent. In one aspect, the method comprises (a) determining a first concentration of melanoma inhibitory activity protein (MIA) in a first biological sample taken from the mammalian subject before treatment with the melanoma inhibitory agent; (b) determining a second concentration of MIA in a second biological sample from the mammalian subject taken after treatment with the melanoma inhibitory agent; and (c) comparing the first and second concentrations of MIA, wherein a decrease in the second concentration of MIA measured in the second biological sample as compared to the first concentration of MIA measured in the first biological sample indicates a positive response to the treatment with the melanoma inhibitory agent.