Evaluation and prediction methods for heat exchangers have significantly developed over the years where the equal flow Reynolds number, equal flow velocity, and Wilson plot techniques are commonly used. However, equal flow Re and equal flow velocity techniques impose restrictions on the flow conditions and the flow passage configuration. Similarly, Wilson plot technique is applicable under a known exponent of Re. Using these techniques, it is difficult to obtain generalized correlations even for single passage because many unknowns are needed to be solved. Therefore, a simplified and handy method to develop heat transfer correlation is required to design new equipment or predict their performance. A general variance minimization method is presented to avoid the need for determination of various coefficients and exponents used in regression process. This method is validated by comparing its results with the experimental data. (C) 2015 Elsevier Ltd. All rights reserved.