Abstract: The purpose of this quasi-experimental research study was to analyze the effects that Directed Reading-Thinking Activity (DR-TA) had on third grade students' reading comprehension in rural school district. The results of the study are presented in the form of demographic information and data analysis for two hypotheses. An analysis of covariance (ANCOVA) and Step-Wise Liner Regression statistical tests were processed by using an alpha level of .05. The sample for this study was selected from ten third grade classrooms in three elementary schools from three rural school districts. School one and two are in the treatment group, and they both are reading first schools. The third school was in the control group, and it was a non-reading first school. There were eighty students (N=80) in the study. There were fifty-one students (n=51) in the control group (traditional reading strategy), and twenty-nine students (n=29) in the treatment group (DR-TA (alternative) reading strategy). Hypothesis 1 stated that there would be no significant difference in third grade students reading comprehension (GMRT posttest) by type of reading strategy (traditional vs. DR-TA) when controlling for pretest reading comprehension. Data analysis generated an F = 3.503, p =.065. The students who had the DR-TA reading strategy did not show significant differences on the GMRT reading test than the students who had the traditional reading strategy, when controlling for GMRT pretest. Thus, the null hypothesis failed to be rejected. Hypothesis 2 stated that there would be no significant relationship between third grade posttest reading scores and a group of predictor variables including type of reading strategy, pretest reading score, students' pre/post MRQ scores on reading efficacy, reading challenge, reading curiosity, aesthetic enjoyment of reading, importance of reading, compliance, reading recognition, reading for grades, social reasons for reading, reading competition, reading work avoidance; pre/post test RAI scores on school reading and reading for your own enjoyment; and teachers' pre/post TORP scores. SPSS led to four models that displayed independent variables that added significantly to the regression equation; either alone or in combination with other variables. The predictor variables were gmpre, curpre, socpost, and grapost.