White matter tractography is a noninvasive method for estimating and visualizing the white matter connectivity patterns in the human brain using diffusion tensor imaging (DTI) data. Sources of experimental noise may induce errors in the measured fiber directions, which will reduce the accuracy of the estimated white matter trajectories. In this study, a statistical nonparametric bootstrap method is described for estimating the dispersion and other errors in white matter tractography results. Prior studies have derived models of tractography error using the signal-to-noise ratio (SNR) and diffusion anisotropy of the DTI data. Tractography errors measured using bootstrap methods were generally consistent with an analytic model of tractography error except in areas where branching was evident. White matter tractography with bootstrap resampling is also applied to estimate the probabilities of connection between brain regions. The approach was used to generate probabilistic connectivity maps between the cerebral peduncles and specific cortical regions.