Asset health monitoring has traditionally relied on data from inspections, tests, sensors and monitoring systems. However, the use of digital models has increased. Within a digital twin (DT) framework, digital models of the most relevant degradation mechanisms can be used to estimate an asset's current health and to predict its future status. A DT can provide support for decision-making aiming at optimizing inspection planning, prioritizing the most critical assets and evaluating relevant parameters to health management, such as assets remaining useful lives (RUL) and inspection timing and spots [1]. From a DT perspective, all information is underpinned by data and digital models [2]. Therefore, data and model uncertainties play a very important role in model estimates and predictions. These uncertainties are related to several issues, such as data quality, process variations, asset design specifications, fabrication tolerances, measurement uncertainties and the representativeness of the degradation model [1,3]. This paper addresses how these uncertainties influence the RUL assessment and impact the reliability of inspection planning for model-based health monitoring. Focusing on O&G subsea assets, the influence of uncertainties on models’ estimates is addressed, as well as its impact on the evaluation of assets RUL, which supports maintenance and inspection planning. This paper emphasizes the importance of considering uncertainties when using DTs for decision-making on asset health monitoring.