In this paper, we propose and evaluate a new conscious eyeblink differentiation method, comprising an algorithm that takes into account differences in individuals, for use in a prospective eyeblink user interface. The proposed method uses a frame-splitting technique that improves the time resolution by splitting a single interlaced image into two fields—even and odd. Measuring eyeblinks with sufficient accuracy using a conventional NTSC video camera (30 fps) is difficult. However, the proposed method uses eyeblink amplitude as well as eyeblink duration as distinction thresholds. Further, the algorithm automatically differentiates eyeblinks by considering individual differences and selecting a large parameter of significance in each user. The results of evaluation experiments conducted using 30 subjects indicate that the proposed method automatically differentiates conscious eyeblinks with an accuracy rate of 83.6 % on average. These results indicate that automatic differentiation of conscious eyeblinks using a conventional video camera incorporated with our proposed method is feasible.