We propose a phenomenological model for high-Tc superconductors (HTSC) assuming: (1) a microscopic phase separation between superconducting and normal-metal areas in the overdoped region; and (2) existence of a homogeneous superconducting phase only below the pseudo-gap T * line, which shows a sharp reduction towards T * ∼ 0 at a mildly overdoped critical concentration xc. This model explains anomalous doping and temperature dependences of ns/m * (superconducting carrier density / effective mass) observed in several overdoped HTSC systems. We point out an analogy to superfluid 4 He/ 3 He films, and discuss an energetic origin of microscopic phase separation. 74.25.Dw, Since the discovery of high-T c cuprate superconductors (HTSC), accumulated studies have revealed unusual phenomena in the underdoped (UD) region, such as, correlations between T c and n s /m * (superconducting carrier density / effective mass) at T → 0 shown by measurements of the magnetic field penetration depth λ [1], and the pseudo-gap phenomena [2]. These results stimulated development of various theories / models for condensation, including Bose-Einstein (BE) to BCS crossover [3,4], phase fluctuations [5,6], XY-model [7] and BE condensation [8]. These models [3][4][5][6][7][8] assume the existence of pre-formed charge pairs above T c . Different pictures, however, assume single charge above T c based on the resonating-valence-bond concepts [9,10].Several anomalous results have been found also in the overdoped (OD) region: µSR studies on Tl 2 Ba 2 CuO 6+δ (Tl2201) [11,12] revealed that n s /m * at T → 0 decreases with increasing hole doping. This tendency has been observed subsequently in thin film La 2−x Sr x CuO 4 (LSCO) [13] and bulk YBa 2 Cu 3 O y (YBCO) [14] systems in the OD region. The "coherence peak" in ARPES spectra [15] also follows this behavior of n s /m * in the OD region. Meanwhile, Tallon and Loram [16] noticed a sharp reduction of the pseudo-gap T * line in the temperature T vs. hole-concentration x phase diagram heading towards T * ∼ 0 at a critical concentration x c ∼ 0.19 holes/Cu which lies in the mildly OD region. In view of the existence of superconductivity in x > x c where the T * line does not exist, Tallon and Loram [16] have advocated a view point that the pseudogap phenomena is not representing pre-cursor superconducting phenomena, but should be ascribed to a tendency towards a ground state competing with superconductivity.We have, however, suggested another possible view point in which T * is ascribed to a signature of a gradual pre-formation of pairs [4] while the anomaly in the OD region to a phase separation [4,11,17,18]. In the present paper, resorting to a model-calculation of n s /m * , comparison with experimental results, a crude estimate of competing electrostatic and condensation/pairing energies, and analogy to 4 He/ 3 He films, we demonstrate that our picture with microscopic phase separation between superconducting and normal-metal regions can quantitatively account for several anomalous results in ...