The work aims to propose a reliability analysis model based on Wiener process and Bayesian method to study the life reliability of corroded submarine pipelines. The corrosion deterioration model of submarine pipelines was established based on the corrosion data of inner wall and wiener process. For different periods of corrosion data, the normal-inverse gamma distribution function was used as the prior distribution for the unknown parameters of the model. The unknown parameters were calculated by an optimized Bayesian method recursively to finally analyze the reliability of corroded submarine pipelines. A corroded submarine pipeline was taken as an example to verify the model. The reliability analysis showed that the corroded pipeline was almost completely reliable the in the first 12 years and the reliability was 99.22%. And 13 years later, the corrosion rate began to accelerate. When the reliability was 90%, the pipeline could be operated for 13.6 years. When the pipeline operated for 15 years, the reliability was 68.97%, and for 20 years, the reliability was 2.4%. When the pipeline operation time reached to the design life of 25 years, the pipeline reliability was almost 0%. The average residual life of the pipeline was 15.99 years. Based on the corrosion data of submarine pipelines, an intuitive analysis model is provided for the reliability of long-life and small-sample submarine pipelines. The analysis model can calculate the corrosion data measured in different times, and with the increase of pipeline operation time, the reliability analysis results can be updated continuously. It does not need to calculate the historical data repeatedly and can be used for real-time reliability analysis of corroded submarine pipelines.