The goal of this study was to evaluate the assessment criteria of learning in tertiary education of Bangladesh. The study was aimed at identifying percentage of questions during learning assessment of business graduates following cognitive levels of Bloom's in revised Taxonomy and to test whether there was any significant difference in the assessment procedure of different semesters. It was also intended to justify the proportion between higher order learning and lower level learning assessment in tertiary education. It was a descriptive research using secondary sources of data in the form of printed questions. Around 350 printed questions of different semesters in the sample business department of a private university in Bangladesh have been analyzed. Microsoft Office Excel 2007 package has been used as a tool of descriptive statistics and calculating Chi-Square Test value. The results reveal that the grand mean percentage in Apply level is the highest (26.10%) and Create level shows the lowest (2.80%). Chi-square test shows that the questions in different semesters are consistent as critical value is higher than the calculated value. Besides, grand mean percentage in higher order learning level (59.04%) is fairly greater than lower level (40.96%) which represents moderated upward trend. Consequently, the study indicates that the performance of the sample business department in assessment of learning of the students following revised Taxonomy is fairly satisfactory and recommends that emphasize should be given more on higher order learning levels. The results of the study would help the higher education & research authorities especially in Bangladesh with the relevant scenario of assessment trend in business education and take necessary steps in facilitating academic excellence.