In Pakistan, mathematics is taught as one of the compulsory subject in all public and private schoolsfrom primary to secondary level. But unfortunately, this subject is not of much interestfor thePakistani students studying in public and private schools; it is rather a nightmare for them. There may be many reasons for this effect but among one of them is the teaching method. Problem solving method is a source of developing problem solving ability, through which students could be able to solve daily life problems. Most of the mathematicians were inspired by the classical work of Polya (1981) and Dewey (1933). Problem Solving Method is helpful in developing Higher Order Thinking Skills (HOTS). According to Anderson (2001), to measure these skills, the Revised Bloom’s Taxonomy is the best evaluation source.The intent of this true experimental study was to find out the effect of George Polya’s problem solving method of teaching on revised Bloom’s Taxonomy of Educational Objectives, in the subject of mathematics at elementary level. The objectives of the study were i) To find out the effect of problem solving method on conceptual knowledge sub-level remembering. ii) To evaluate the effect of problem solving method on conceptual knowledge sub-level understanding. iii) To check the effect of problem solving method on conceptual knowledge sub-level applying. iv) To determine the effect of problem solving method on conceptual knowledge sub-level analyzing. v) To find out the effect of problem solving method on conceptual knowledge sub-level evaluating.vi) To check the effect of problem solving method on conceptual knowledge sub-level creating. vii)To find out the effect of problem solving method 10 on Higher Order Thinking skills of Revised Bloom’s Taxonomy. To achieve the objectives, eight research questions were developed. To verify the objectives, eight null hypotheses were developed. True experimental pretest-posttest (double control group) design was used.An experiment was performed in a school situated in urban area of Islamabad. Multistage sampling technique used for the selection of school. 132 8th grade students were used in the experiment. Pre-test which consisted of 60 items of six cognitive processes was developed. Validity of the test was checked by the experts and reliability was checked throughAlpha reliability analysis which wasα = .89. On the basis of pre-test scores, three groups were formed through proportionate random sampling. The Experimental Group was taught by the researcher through Problem Solving Method, and other two groups were taught through Conventional Method by the same school teachers. All possible efforts were made to control the internal and external threats. Forty lessons were taught during the 8-week experiment to all three groups. Post-test was developed by changing the order of the questions of pre-test. Post-test was administered and marked according to the rubric design for marking. Pre-test and post-test scores were analyzed on SPSS. One Way ANOVA and t-test were applied on the data.In the light of the analyses of data it was found that by teaching through Problem Solving Method, students performed better on Revised Bloom Taxonomy as compared to Conventional method. It was concluded on the bases of findings that Problem Solving Method showed significant improvement at all levels of knowledge dimension and hence all null hypotheses were not accepted. It was also concluded that Problem Solving Method works better than Conventional Method for teaching of Mathematics. In the light of these conclusions, Problem Solving Method was recommended for teachingof Mathematics at Elementary level and also suggested to add it in teachers training programmes as well.