The recent research aimed to identify the effect of differences in items' order in an achievement test of "learning psychology subject" on students' academic performance and (difficulty, discrimination, and guessing) coefficients of test items according to Three Parameter Logistic Model depending on Item Response Theory. Two models of an achievement test in "learning psychology subject" were prepared; the first (A) was ordered according to content’s order, the other (B) was ordered randomly. The final version of the test contained (70) items; (50) multiple choice items, and (20) true or false items. The final sample was (2033) male and female students enrolled in general diploma in Education –the first semester academic year 2018/2019, and it was corrected automatically via scanner. Suitable statistical processes were used to ensure the accordance of final sample data with Item Response Theory hypotheses. Using (Bilog-Mg V3.0) these parameters (difficulty, discrimination, and guessing coefficients) and standard error in estimating each of them according to Three Parameter Logistic Model were calculated, also the effect of model type (A) and (B) on academic achievement using independent samples T-test were studied. Results showed that: There are no statistically significant differences in estimations of difficulty, discrimination, and guessing coefficients and their standard errors according to model type (A) and (B), and there are no statistically significant differences in students' academic performance according to model type (A), and (B).