Comparison of blood and feather pulp (FP) samples for the diagnosis of Marek's disease (MD) and for monitoring Marek's diseases vaccination in chickens (serotypes 2 and 3 vaccines) by real time-PCR was evaluated. For diagnosis of MD, quantification of serotype 1 Marek's disease virus (MDV) DNA load was evaluated in 21 chickens suffering from MD. For each chicken, samples of blood and FP were collected and MDV DNA load was quantified. Solid tumors are the sample of choice for MD diagnosis by real time-PCR and, hence, 14 solid tumors were included in the study as positive controls. Load of MDV DNA in FP was equivalent to that detected in solid tumors (threshold cycle [Ct] ratio above 1.7). MDV DNA load in blood samples was lower than in solid tumors and FP samples. Nonetheless, there was a statistically significant correlation of the results obtained from FP and blood (r  =  0.92). Results of the Pearson correlation test showed that Ct ratio values of 1.7 in FP correspond to Ct ratio values of 1.2 in peripheral blood. For monitoring vaccines, serotypes 2 and 3 MDV DNA load was evaluated in blood and FP samples of vaccinated chickens. Serotype 2 MDV DNA load was evaluated in samples of blood and FP from 34 chickens vaccinated with SB-1 strain. Serotype 3 MDV DNA load was evaluated in blood and FP samples from 53 chickens vaccinated with HVT strain. For both serotypes, frequency of positive samples and load of vaccine DNA was higher in FP than in blood samples. There was not a statistically significant correlation between the load of SB-1 DNA (r  =  0.17) or HVT DNA (r  =  −0.04) in FP and blood. Our results show that the load of serotypes 1, 2, and 3 DNA is higher in FP than in blood. Diagnosis of MD could be done using both FP and blood samples. Monitoring of MD vaccination by real time-PCR required the use of FP samples. There were a high percentage of false negative samples when using blood to detect serotypes 2 and 3 MDV by real time-PCR.