We have performed LC-MS based plasma proteomics of twenty febrile patients coming in at Peijas Hospital, Helsinki University Hospital (Vantaa, Finland). At the same time, blood culture was performed to classify the patients to blood culture positive and negative for bloodstream infections. Blood culture negative patients served as controls for the positive ones and these two groups were used for identifying protein profile changes occurring in positive patients. We have subsequently performed various statistical analyses such as principal component analysis as well as pathway analysis by two different methods. We show, in the current study, proteins significantly differing in these two groups as well as dysregulated pathways. These changes seem not to depend on bacterial species as different patients had different blood stream infections. Further, we statistically analyzed the proteomic dataset to find the potential biomarkers which can classify the two classes of patients.