In this paper, the validity of different approximations used in the calculation of induced overvoltages in power lines are investigated, These approximations are as follows: 1) neglect the distortions introduced by the finitely conducting ground on the electromagnetic (EM) fields; 2) the horizontal electric field at ground level is calculated by using the wavetilt approximation, which is valid for radiation fields and for grazing incidence: 3) the horizontal held at line height is obtained by adding the horizontal held calculated at ground level to the horizontal field at line height calculated over perfectly conducting ground; 4) the transmission line equations derived by assuming that the ground is perfectly conducting are used with the horizontal field present over finitely conducting ground as a source term in calculating the induced overvoltages; and 5) the propagation effects an the transients as they propagate along the line are either neglected or modeled by replacing the line impedance due to ground by a constant resistance, The results presented in this paper show that in the calculation of induced overvoltages the approximation 3) is justified and approximation 2) is justified if the interest is to estimate the peak value of the induced overvoltage, Approximation 4); is probably justified for short lines and/or for highly conducting grounds. But it can introduce significant errors if the line is long and ground conductivity is low. Approximations 1) and 5) may lead to significant errors in the peak value, risetime, and derivative of the lightning induced overvoltages.