Several publications have suggested using spectral reflectance indices to monitor grapevine water status to replace the traditional pressure chamber technique. But these indices do not always provide good results and require some adjustments. For this reason, optimization of spectral reflectance indices is proposed for improving the estimation of the midday stem water potential (Ψstem) in grapevines. Thus, a two-season trial was established in a commercial vineyard (Pencahue Valley, Chile) under two irrigation levels implemented between fruit set and veraison. Optimization of 300 spectral indices was carried out using a contour map methodology obtained by the authors computer program (ANDES). Results indicated that the spectral reflectance indices in the visible bands of the spectrum presented the best correlations with (Ψstem). In this case, the best coefficients of determination (R2) values after the optimization process were obtained with BI Index (BI), Simple Ratio 4 (SR4), normalized pigments chlorophyll index (NPCI) and red green index 2 (RGI2) indices. Furthermore, the optimization process improved the estimation of Ψstem by 33, 11.2, 48.1 and 21%, respectively.