Image formation in multiphoton fluorescence microscopy through double-layer turbid tissue media is investigated using Monte Carlo simulation. With the help of the concept of the effective point spread function, the relationship of image resolution and signal level to the thickness and scattering properties of the double-layer turbid media under single-, two-, and three-photon excitation is revealed. Results show that for a double-layer turbid medium of a given thickness, small particles in the top layer result in a quicker degradation of signal level than large particles in the top layer. This model is then applied to study the penetration depth of multiphoton fluorescence microscopy through human skin tissue which exhibits a layered structure. It is predicated that using 3p excitation leads to a signal level up to two orders of magnitude higher than that under 2p excitation, while diffraction-limited image resolution can be maintained for skin tissue of thickness up to 500 µm.