The paper addresses a method for the assessment of variances of responses at transmission lines (TL) with randomly varying primary parameters, based on the theory of stochastic differential equations (SDE). In contrast to previous works where stochastic trajectories were statistically processed to get sample means and sample standard deviations, herein the statistical characteristics are solved directly through ordinary differential equations (ODE) prepared. The TL model is formed by cascade connection of the RLGC-based Π-sections enabling to model also nonuniform TLs if necessary. To develop the model equations a state-variable method is applied, and then a corresponding vector SDE is formulated. To verify the novel approach the SDEs are solved to obtain batches of individual stochastic trajectories, and then sample statistics and corresponding confidence intervals are computed. A stochastic implicit Euler numerical scheme is used for this purpose while utilizing the Matlab language.