The learning curve is one of the tools in operation planning and control. It enables the planner to estimate the future performance of tasks executed either by individuals or organizations. Choosing the appropriate mathematical model that depicts learning is usually based on a “least square oriented” searching procedure, hence prior knowledge of the possible model that may fit a certain operation would facilitate the searching procedure. This study investigates the influence of three groups of variables on the chosen model: working conditions, type of instructions, and previous experience. The following three models were investigated: linear, power, and exponential. Results show that the predictability power of the model is a function of operational variables as well as the specific model which is applied.