The purpose of this study was to improve the existing short-form (SF) of the K-WAIS-IV to be used in the Korean military draft examination. The existing SF of the K-WAIS-IV (Choi et al, 2014), when used to the subnormal group, overestimated the IQ of the subnormal group because of regression toward mean. Also, the existing SF’s subscales were not enough to measure fluid reasoning (Gf) component. To address these problems, this study suggests three new SFs: SF A, SF B, and SF C. SF A (including VC, AR, MR, and CD) and SF B (VC, AR) were developed based on the confirmation factor analysis (CFA) result of draft examination sample (n=474). SF C (AR, MR), with theoretical and practical reasoning that evaluation of Gf would be useful in draft examination, includes arithmetic reasoning (AR) and matrix reasoning (MR). A multiple regression analysis was performed to derive a regression formula equation to estimate IQ. All three SFs met all validity criteria (Resnick & entin, 1971) with the exception of concordance rate. In an additional study conducted to examine utility of the new SFs as a screening test, SF A showed a very high screening ability. SF B and SF C also showed a decent screening ability. Interestingly, SF C was shown to be better than SF B in selecting out those who need an additional in-depth examination. The ROC results indicated that SF A, SF B, and SF C are all secondary screening test. Finally, implications and limitations of this study as well as suggestions for further research were discussed.