Based on evidence from morphological and histochemical studies and from clinical experience, the following hypotheses are proposed: carcinoma-in-situ (CIS) germ cells are malignant gonocytes; these CIS gonocytes have some capacity to regress into more primitive, totipotent embryonic cells which can give rise to all types of nonseminomatous germ cell tumours; the tumour germ cells of classical seminomas are malignant gonocytes derived from CIS gonocytes which have lost their ability to regress into totipotent embryonic cells; the ability of CIS gonocytes to regress into totipotent embryonic cells decreases with age, whereas the capacity to form classical seminoma cells is preserved; the transformation of CIS gonocytes into invasive tumours is dependent on factors such as gonadotrophins and/or testicular steroids; the pathogenesis of classical and spermatocytic seminoma are unrelated. As a consequence of these hypotheses an alternative nomenclature for carcinoma-in-situ, seminoma and dysgerminoma is suggested.