The controlled-source audio frequency magnetotelluric (CSAMT) method is well established in geophysical prospecting. By using a transmitter (usually a grounded electric dipole) to generate the source fields, the CSAMT method overcomes problems associated with weak natural EM fields used in conventional audio MT surveys. But the CSAMT method creates a new problem associated with active EM surveys: when anomalies are located between the transmitter and receiver, the measured data may be distorted so that the geological structure will be inferred incorrectly. Kuznetzov (1982) and Zonge and Hughes (1988) observe this phenomenon in 2D numerical model studies and term such behavior the shadow effect. This phenomenon has been simulated by numerical modeling and discussed in many papers (Newman et al., 1986; Zonge et al., 1986; Mitsuhata, 2000). Boschetto and Hohmann (1991) use numerical methods to study shadow and source overprint effects in CSAMT using a horizontal electric dipole source. For this problem, we separate the ground-wave and surface-wave terms from the closed-form expressions of the electric and magnetic fields in a uniform earth and use the ground-wave/surface-wave ratio to estimate the potential magnitude for producing shadow and source overprint effects. This is a new and effective analytical method.