<p>Tractography is a unique modality for the in vivo measurement of structural connectivity, crucial for understanding brain networks and neurological conditions. With increasing <i>b</i>-value, the diffusion-weighting signal becomes primarily sensitive to the intra-axonal signal. However, it remains unclear how tractography is affected by this observation. Here, using open-source datasets, we showed that at high <i>b</i>-values, DWI reduces the uncertainty in estimating fiber orientations. Specifically, we found the ratio of biologically-meaningful longer-range connections increases, accompanied with downstream impact of redistribution of connectome and network metrics. However, when going beyond <i>b</i> = 6000&#xa0;s/mm², the loss of SNR imposed a penalty. Lastly, we showed that the data reaches satisfactory reproducibility with <i>b</i>-values above 1200&#xa0;s/mm². Overall, the results suggest that using <i>b</i>-values above 2500&#xa0;s/mm² is essential for more accurate connectome reconstruction by reducing uncertainty in fiber orientation estimation, supporting the use of higher <i>b</i>-value protocols in standard diffusion MRI scans and pipelines.</p>