In the field of thermal engineering, one of the biggest concerns is the cooling of heat producing systems. For this purpose, today’s world is encouraging to use such cooling systems which are free from any active components (passive systems) for its high reliability and compact size. For this reason, to establishcooling by transferring heat from one place (source) to another (sink) passive system like natural circulation loop (NCL) is highly used. Fluid flow dynamics of the NCL is changing with the increase in heater power which is used as the source for the simulation. We found steady flow dynamics for the comparatively low power of heat, and with the rise in the power first, we saw the oscillatory flow dynamics and then found flow reversal characteristics. This paper presents a novel strategy for the early prediction of flow reversal phenomenon in NCLusing symbolic analysis of time series data. This time series data is found from the numerical simulation, and for the proper study, we are considering data after the initial transient part is overcome. Total time series data is transformed into a symbol string by partitioning into a finite number of specified symbolised groups. The state probability vector is calculated based on the number of occurrences of each symbol group. Present work is a single-phase study, and according to our geometry, we can provide a maximum 800 W heater power to stay in the single-phase. Therefore, for the early prediction of flow reversal in NCL, state probability vector evaluated at 800 W heater power which is the most undesirable state (chaotic data), and this is considered as the reference vector. The difference of the reference state vector from the current state vector is used as a parameter for early detection of flow reversal. It can be observed from the results that this difference changes significantly when the system is sufficiently away from the flow reversal.