The yellow fever 17D virus is attenuated and used for human vaccination. Two of its substrains, 17D-204 and 17DD, are used for vaccine production. One of the remarkable properties of this vaccine is limited viral replication in the host but with significant dissemination of the viral mass, yielding a robust and long-lived neutralizing antibody response. The vaccine has excellent records of efficacy and safety and is cheap, used as a single dose, and there are well-established production methodology and quality control procedures which include the monkey neurovirulence test (MNTV). The present study aims at a better understanding of YF 17DD virus attenuation and immunogenicity in the MNVT with special emphasis on viremia, seroconversion, clinical and histological lesions scores, and their intrinsic variability across the tests. Several MNVTs were performed using the secondary seed lot virus 17DD 102/84 totaling 49 rhesus monkeys. Viremia was never higher than the accepted limits established in international requirements, and high levels of neutralizing antibodies were observed in all animals. None of the animals showed visceral lesions. We found that the clinical scores for the same virus varied widely across the tests. There was a direct correlation between the clinical scores in animals with clinical signs of encephalitis and a higher degree of central nervous system (CNS) histological lesions, with an increase of lesions in areas of the CNS such as the substantia nigra, nucleus caudatus, intumescentia cervicalis, and intumescentia ventralis. The histological scores were shown to be less prone to individual variations and had a more homogeneous value distribution among the tests. Since 17DD 102/84 seed virus has been used for human vaccine production and immunization for 16 years with the vaccine being safe and efficacious, it demonstrates that the observed variations across the MNVTs do not influence its effect on humans.