Recently, animal testing has been affected by increasing ethical, social, and political concerns regarding animal welfare. Several in vitro safety tests for evaluating skin sensitization, such as the human cell line activation test (h-CLAT), have been proposed. However, similar to other tests, the h-CLAT has produced false-negative results, including in tests for acid anhydride and water-insoluble chemicals. In a previous study, we demonstrated that the cause of false-negative results from phthalic anhydride was hydrolysis by an aqueous vehicle, with IL-8 release from THP-1 cells, and that short-time exposure to liquid paraffin (LP) dispersion medium could reduce false-negative results from acid anhydrides. In the present study, we modified the h-CLAT by applying this exposure method. We found that the modified h-CLAT is a promising method for reducing false-negative results obtained from acid anhydrides and chemicals with octanol-water partition coefficients (LogK) greater than 3.5. Based on the outcomes from the present study, a combination of the original and the modified h-CLAT is suggested for reducing false-negative results. Notably, the combination method provided a sensitivity of 95% (overall chemicals) or 93% (chemicals with LogK > 2.0), and an accuracy of 88% (overall chemicals) or 81% (chemicals with LogK > 2.0). We found that the combined method is a promising evaluation scheme for reducing false-negative results seen in existing in vitro skin-sensitization tests. In the future, we expect a combination of original and modified h-CLAT to be applied in a newly developed in vitro test for evaluating skin sensitization.