Myelin facilitates efficient neural signal transmission along axons. While predominantly present in white matter, myelin is also found in smaller amounts in the outermost layer of gray matter, known as the cortex, as well as deep subcortical structures. Quantitative magnetic resonance imaging (qMRI) metrics, including longitudinal relaxation rate (R1), can report myelin levels in vivo, although these measures are often estimated and can vary depending on pulse sequences and acquisition parameters used to collect the data. Multi-site approaches can increase the statistical power and ecological validity of studies by recruiting participants from larger and more diverse samples, but considerable variability has previously been reported in qMRI estimates of myelin computed using data from different MRI scanners. Here, we assessed intra- and inter-vendor variability in R1 maps acquired at three sites using scanners from two manufacturers, General Electric (GE) and Siemens. Two ‘traveling heads’ scans, where a participant visits and completes a scan at each site, were acquired to directly assess site effects independent of biological variability. Histogram-based site-specific scaling factors determined from the ‘traveling heads’ scans were then applied to harmonize the main dataset. Before site correction, comparable inter- and intra-vendor differences were observed in both the main dataset and the ‘traveling heads’ scans. Our method of harmonization reduced the inter-site variability considerably, while remaining sensitive to age effects in R1. We then characterized age trajectories of R1 across the whole brain using linear regressions. In the age range of 16-36, age was significantly associated with mean R1s of bilateral putamen, left pallidum, and left amygdala, in line with past research. Cortical analyses also replicated earlier findings, with higher age associations in regions of the motor and lateral frontal cortices. The findings contribute to the growing research in multi-site harmonization, while providing normative estimates and trajectories of whole-brain R1s that can be referred to in future clinical studies. / Thesis / Master of Science (MSc) / Myelin is crucial for efficient signal transmission in nerve cells, and advancements in magnetic resonance imaging (MRI) have allowed us to indirectly estimate myelin levels in the brain. Collaborative multi-site investigations have become increasingly popular due to their ability to recruit larger and more diverse samples, making their findings more generalizable to the public. However, differences in scanner hardware and configurations across sites can introduce site-dependent biases to MRI scans. In this thesis, we found considerable inter-site variability in our whole-brain MRI-based estimate of myelin based on longitudinal relaxation rate (R1). We thus developed a method that can remove unwanted site-dependent differences from images by comparing repeated scans across sites. Using the corrected data, we showed how myelin amounts in different brain structures changes with age during early adulthood. Overall, we highlight the importance of carefully considering site-specific differences in multi-site MRI data and characterized how myelin changes with age in young adults.