Events related to the very early steps of steroid hormone action are reviewed, after a brief exposé of the methods used for the measurement and calculation of protein-steroid interactions. A list of steroid binding plasma proteins is given, and their physiological variations and possible role(s) are discussed. The mechanism of entry of steroids into target cells is then considered especially the possibility of a protein mediated step. The metabolism of steroid hormone in target cells and its physiological significance are discussed, especially with reference to androgens in prostate, levator ani muscle and hypothalamus. No receptor of steroid hormone has yet been discovered if defined as the last entity with which the active steroid interacts before action is initiated. However, in the target tissues of steroid hormones, specific proteins have been detected (termed receptors) with properties that are compatible with an essential if not an obligatory role in steroid action. The properties ascribed to the cytosol receptors of different steroid hormones in various tissues, their binding properties, and their quantitative variations are reported. The transformation or activation of receptor after binding of the hormone is related to its transfer to the nucleus. The question posed by the recent observations on Non Histone Chromatin estrogen binding protein and of the "insoluble" nuclear receptor are discussed. Finally, a series of experiments are reviewed, dealing more specifically with the increase of protein and RNA synthesis in tissues in which steroids and RNA synthesis in tissues in which steroids promote growth (i.e. estradiol in uterus). The concept of a cascade phenomenon is reported, including the early synthesis of KIP (Key Intermediary Protein(s)) which would responsible for the secondary amplification of the response. It is discussed in connection with the present (and insufficient) knowledge of receptors and transcription machinery.