It is a goal of cancer researchers to identify the most aggressive subpopulations of tumour cells for the development of novel therapeutics for patients with advanced stage disease. However, current approaches to identify and isolate these tumour-initiating subpopulations of tumour cells are extremely varied, and this has limited the ability of researchers to determine universal markers of such subpopulations for a variety of cancers. Cutaneous melanoma is an example of a malignancy where there is much debate over the identity of tumour-initiating cancer cells and there is a lack of well-defined cell surface markers to consistently identify these subpopulations in tumours. It is becoming increasingly apparent that tumour-initiating cells are not a fixed population, but rather their tumour-initiating phenotype can be heavily influenced by the tumour microenvironment. Growing melanoma cells in 3D culturing conditions provides an artificial environment that can select for tumour-initiating phenotypes. Here we shall circumvent the requirement of robust markers for isolating tumour-initiating subpopulations from genetically-diverse melanomas, by culturing cells in 3D conditions. Cells isolated from these conditions will be used to investigate the behaviour of these cells and identify novel biomarkers of highly aggressive subpopulations of tumour cells.