The jet breakup length is one of important phenomenological parameters to evaluate the corium debris bed coolability during the progress of severe accident in nuclear power plant, since the parameter is related to the determination of the liquid melt fraction in debris bed. Existing jet breakup length data reveal relatively large scattering and different trends, similar to the Saito or Epstein correlation, depending on cases. The scattering and separation of the data may be due to uncontrolled factors in the melt discharge condition and also inconsistent evaluation methods used for the jet breakup length.Through melt jet breakup experiments with low melting point metal alloy (Bi-Sn alloy), the influence of the estimation methods for the jet diameter (measured or calculated value) and the jet breakup length (position- or velocity-based value) was investigated. According to the comparison between the measured and calculated jet diameters, the experimental uncertainty was reduced by producing well-controlled melt jet. The concept of the jet breakup length was revisited in detail, emphasizing the shorter breakup length by the velocity-based method. The modification of the Saito correlation constant from 2.1 to 1.7 was suggested to fit the velocity-based breakup length. The importance of a well-defined data analysis process for non-dimensional jet breakup length was highlighted for the minimization of the experimental uncertainty.