Resistance as a function of temperature for optimally doped manganese perovskites is simulated based on a random resistor network model which is generated through the Monte Carlo method by identifying neighbours in each cubic lattice randomly occupied by ferromagnetic metallic particles. Only one parameter, i.e. the number density of ferromagnetic particles, is used as an adjustable parameter, and it is shown that the model yields excellent agreement with measured resistance data in (La1-xYx)2/aCa1/aMnO3 (0 ≤ x ≤ 0.2) for temperatures covering the whole range from the high-T insulating bebaviour to the low-T metallic state.