The disease leptospirosis is known to be endemic in Malaysia, and it significantly impacts human wellbeing and the national economy. Current surveillance systems are based on morbidity and mortality leptospirosis national data from the Ministry of Health and remain inadequate due to the number of unreported and misdiagnosed cases. A robust surveillance system is needed to monitor temporal and spatial changes which yield improvements in terms of identifying high-risk areas and disease behaviour. The objective of this study is to identify high-risk areas by estimating relative risk using existing models which are the Standardized Morbidity Ratio (SMR), Poisson-gamma, log-normal, Besag, York and Mollié (BYM) and mixture models. An alternative model is also proposed which involves transmission systems and stochastic elements, namely the stochastic Susceptible-Infected-Removed (SIR) transmission model. This estimation of risk is expected to assist in the early detection of high-risk areas which can be applied as a strategy for preventive and control measures. The methodology in this paper applies relative risk estimates to determine the infection risk for all states in Malaysia based on monthly data from 2011 to 2018 using WinBUGS 1.4 software. The results of relative risks are discussed and presented in tables and graphs for each model to disclose high-risk areas across the country. Based on the risk estimates, different models used have different risk interpretations and drawbacks which make each model different in its use depending on the objectives of the study. As a result, the deviance information criteria (DIC) values obtained do not differ greatly from each expected risk which was estimated.