A systematic evaluation of 768 data-processing pipelines for functional connectomics from resting-state fMRI data reveals significant variability in their performance. The study assesses pipelines based on their ability to minimize motion-related and spurious test-retest discrepancies, while being sensitive to individual differences and experimental effects. The results show that most pipelines fail to meet at least one criterion, but a subset of pipelines consistently satisfy all criteria across different datasets. These optimal pipelines are characterized by the use of weighted edges, Pearson correlation for connectivity, and the OMST filtering method. The study highlights the importance of end-to-end evaluation of pipelines to ensure reliable and reproducible results in functional connectomics. The findings emphasize the need for standardized and validated pipelines to avoid misleading conclusions in neuroimaging research.A systematic evaluation of 768 data-processing pipelines for functional connectomics from resting-state fMRI data reveals significant variability in their performance. The study assesses pipelines based on their ability to minimize motion-related and spurious test-retest discrepancies, while being sensitive to individual differences and experimental effects. The results show that most pipelines fail to meet at least one criterion, but a subset of pipelines consistently satisfy all criteria across different datasets. These optimal pipelines are characterized by the use of weighted edges, Pearson correlation for connectivity, and the OMST filtering method. The study highlights the importance of end-to-end evaluation of pipelines to ensure reliable and reproducible results in functional connectomics. The findings emphasize the need for standardized and validated pipelines to avoid misleading conclusions in neuroimaging research.