The common theme and aim for this thesis was to explore the epidemiology and the diagnostic possibilities of patients with non-malarial febrile infections of tropical origin, with both the individual patient perspective and the more general public health aspects in focus. Analysis from two study areas, Sweden and Mozambique, are presented in the project. The infectious disease panorama in Mozambique has to a large extent been a blind spot. Further epidemiological studies, aiming at more knowledge to guide decisions on preventive measures, are needed. We performed two prospective investigations. The first one was a pilot sero-epidemiological study on vector-bomne viral zoonoses, in which we screened serum samples from patients attending a health care clinic in the suburb of the capital Maputo. In the analysis we found that 29% of the patients screened had an antibody response against one or more of the viral pathogens. Our conclusion, based on these results, was that exposure to chikungunya virus (CHIKV), dengue virus (DENV) and Rift Valley fever virus (RVFV) had taken place, and that these viruses are circulating 1n the country. The second study was an Investigation of the DENV outbreak in the cities Pemba and Nampula. We analysed serum samples for DENV by PCR from patients seeking medical attention for fever during 2015- 2016. The results including PCR positive samples, serotyping and sequencing of strains, confirmed that DENV serotype 2 is now endemic in northem Mozambique. In the Swedish multicentre study we prospectively included febrile travellers returning to Sweden from tropical areas defined as malaria endemic. Epidemiological data from questionnaires and clinical diagnoses given to patients by their attending doctors were first compared with results from a panel of extended laboratory diagnostics, primarily on convalescent samples. We then focused on the possibilities for early diagnostics with PCR, particularly for cases where no relationship between the febrile illness and a microbial pathogen had been identified. We also developed a universal PCR for diagnosis of early phase DENV infection. This universal single probe real time RT-PCR for DENV was then used for the DENV analysis on acute samples. The results showed that infectious disease clinicians in Sweden, when taking care of febrile travellers returning from the tropics, were in general able to establish a diagnosis based on laboratory diagnostics for relevant pathogens. That being said, we also noted that 30 % of the patients included in the study were dismissed with the diagnosis fever of unknown origin. It was also apparent from the results that influenza virus infection was a frequent, and often missed, diagnose among febrile travellers, regardless of the time of year. The antibody screening also identified several additional cases of dengue infection. When including the PCR for DENV in the diagnostic kit, it was possible to reduce even further the number of cases with diagnosis of unknown fever. This confirms that the universal PCR for DENV is a sensitive, specific and valuable diagnostic tool to use during the first 5 days in the acute phase of iliness. Apart from influenza and DENV, Rickettsia and Leptospira infections stood out as differential diagnoses that needed to be addressed.