Determining the separate energy costs of the positive and negative mechanical work in repetitive lifting or lowering is quite complex, as a mixture of both work components will always be involved in the up- and downward motion of the lifter's body mass. In the current study, a new method was tested in which coefficients specifically related to the positive and negative work were estimated by multiple regression on a data set of weight-lifting and weight-lowering tasks. The energy cost was obtained from oxygen uptake measurements. The slopes of the regression lines for energy cost and mechanical work were steeper for positive than for negative work. The cost related to the negative work was approximately 0.3–0.5 times the cost of the positive work. This finding is well in line with data obtained directly from other isolated activities of either positive or negative work (e.g., ladder climbing vs. descending). However, the intercept values of the regression lines were not significantly different from zero or were even negative. This was most likely due to the metabolic energy not related to processes that yield mechanical work (e.g., isometric muscle actions) that was not constant among trials.