In order to explore the differences between measured glycemic index （GI） and glycemic load （GL） with predicted values， test meal A1， A2， B and C， each contains 4 whole grains selected from 9 candidate materials were prepared and their GI were tested in 10 healthy subjects and their GL and hydrolysis index （HI） were calculated. Test meal A2， B， C had GI values of 60， 66 and 48， respectively， and the GL value were 23， 23 and 16. However， the test meal A2 prepared with the default cooking program for whole grains resulted higher GI value of 69 and GL value of 26. Neither the estimated GI value from published GI data nor from the in vitro carbohydrate hydrolysis index gave good prediction on test meals containing pulses.