Abstract: Dengue is today the most important arthropod-borne viral disease and affects over one hundred million people each year. The disease is caused by a RNA virus, which is transmitted by the bite of infected mosquitoes of the genus Aedes, present all over the tropical and subtropical regions of the world. To date, vaccines are not available and the control of the disease depends strongly on the control of the vector and the availability of diagnostic services in the areas of transmission. Currently, the main approaches used for dengue diagnosis are the identification of specific antibodies in the patient's serum by MAC-ELISA (IgM antibody-capture enzymelinked immunosorbent assay) or the detection of viral RNA by reverse transcription - polymerase chain reaction (RT-PCR). The efficiency of those two techniques, however, depends on the quality of the reagents and the protocol utilized. The main objectives of this work were: 1) to obtain specific antigens in a low cost way, able to replace those produced in mouse brain currently used in a MAC-ELISA test developed in Brazil; 2) to establish a protocol for real time PCR to detect dengue virus RNA in human serum samples, utilizing as a prototype the serotype 1 of the virus (DEN1). The dengue antigens for the MAC-ELISA were produced in culture cells and evaluated using the sera of 50 suspected cases of dengue infection. The results were compared to those obtained by the test performed with the original antigen. Only two discordant results were noticed. The analysis of those two samples by a third technique (dot blot) confirmed the results obtained with the antigens produced in cell culture, suggesting that they are more sensitive and specific than the mouse brain extract. Both of the serological tests detected a larger number of positive samples among those collected after the seventh day of the onset of the symptoms. The real time PCR protocol developed in this work was initially evaluated using RNA from dengue virus serotype 1 (DEN1), serotype 2 (DEN2), and yellow fever virus. The sensitivity and specificity of a nested RT-PCR described before were also evaluated in the same manner. The nested PCR was able to detect the RNA originating from 2x102 FFU/ml, while the sensitivity of the real time PCR was lower than 6,25x10"1 FFU/ml. The real time PCR was not able to amplify samples of DEN2, but detected as positive the yellow fever samples; it did not occur with the nested PCR. The same panel of serum used in the ELISA was submitted to the two PCR techniques and the results obtained were compared. The lower sensitivity of the nested PCR was confirmed by the low rate of detection observed when RNAs extracted directly from serum samples were used. However, when the virus were amplified in cell culture prior to RNA extraction, the positive number increased, indicating that such procedure is essential for this technique. The higher sensitivity of the real time PCR seems to compensate, in part, this problem. For the two PCR based techniques the detection of viral RNA were greater among the samples collected before the seventh day after the onset of the disease. These results suggest that the MAC-ELISA should be used in conjunction with an RT-PCR approach in order to improve the efficiency of dengue diagnosis. In addition, the real time PCR seems to be more suitable for the dengue diagnosis since it's more sensitive and easier to perform than the nested PCR.