Leptospirosis is a zoonotic spirochaetal illness that is endemic in many tropical countries. The research base on leptospirosis is not as strong as other tropical infections such as malaria. However, it is a lethal infection that can attack many vital organs in its severe form, leading to multi-organ dysfunction syndrome and death. There are many gaps in knowledge regarding the pathophysiology of leptospirosis and the role of host immunity in causing symptoms. This hinders essential steps in combating disease, such as developing a potential vaccine. Another major problem with leptospirosis is the lack of an easy to perform, accurate diagnostic tests. Many clinicians in resource limited settings resort to clinical judgment in diagnosing leptospirosis. This is unfortunate, as many other diseases such as dengue, hanta virus, rickettsial infections, and even severe bacterial sepsis, can mimic leptospirosis. Another interesting problem is the prediction of disease severity at the onset of the illness. The majority of patients recover from leptospirosis with only a mild febrile illness, while a few others have severe illness with multi-organ failure. Clinical features are poor predictors of potential severity of infection, and therefore the search is on for potential biomarkers that can serve as early warnings for severe disease. This review concentrates on these three important aspects of this neglected tropical disease: diagnostics, developing a vaccine, and potential biomarkers to predict disease severity.