Survival data obtained in the context of several different failure mechanisms are usually analyzed via the theory of competing risks. Competing risks is the subdiscipline of survival analysis where, in addition to the survival time, the cause of the failure is observed. In this study we have describe the differential feature of competing risks data and describe competing risks models in detail both for latent failure time approach and a multistate approach. The main aim of this study is to quantify the risks factors that are associated with particular cause of failures e.g. (coronary heart disease, retinopathy, and nephropathy) for the diabetes mellitus patients. The potential risks factors at the transitions can be identified considering the competing risks as a special case of multistate model. Thus our main aim is to modeling the competing risks using cause-specific hazards. For the analysis purposes the semi-parametric Cox’s proportional hazards model has been treated exactly in a similar manner by considering each mode of failure separately. Considering the search of alternative modeling approach the Weibull regression models fit the data well. Hence, a Weibull regression models is also fit and the results are compare with the results of semi-parametric approach. Our study was based on a set of 1726 type II diabetic patients from BIRDEM who were follow-up during the period 1984 to 1996. For instance, we have found that age has significant effect on each of the failure type. Patients with normal blood pressure has a better survival prognosis than those with high blood pressure. For Body Mass Index (BMI) variable, we have found that overweight patients experienced higher hazard than the ideal BMI condition patients. The results also indicate that, the prevalence rate of Coronary heart disease and Retinopathy are approximately similar and is higher than Kidney complication.