During the past decade new insights have been gained into the role of T lymphocytes in the host's immune response to cancer in general and to melanoma in particular. Several melanoma-associated antigens (MAA) recognized by T cells have been characterized, and a number of HLA class I- and class II-restricted peptides have been identified. These antigens can be divided into three different groups: tumor-associated testis-specific antigens, melanocyte differentiation antigens, and mutated or aberrantly expressed antigens. These proteins give rise to several antigenic peptides. The number of known melanoma-associated peptides that can induce killing by cytotoxic T-lymphocytes (CTL) exceeds 30 and is still increasing. In line with these findings, clinical data indicate that the immune system is essential in the control of tumor growth. A brisk infiltration of lymphocytes is associated with a favorable prognosis, and complete or partial regression of primary melanoma occurs quite frequently. Furthermore, immunomodulatory therapies have accomplished complete or partial tumor regression in a number of patients. However, the immune response is in most cases inadequate to control tumor growth as tumor progression often occurs. Hence, the coexistence of a cellular immune response in melanoma lesions, demonstrated by the presence of clonally expanded T cells, remains a major paradox of tumor immunology. In the present paper we review current knowledge regarding tumor infiltrating lymphocytes (TIL) in melanoma and discuss possible mechanisms of escape from immune surveillance.