There is an understanding that effective HlY / AIDS prevention programs should be based on thernmodels and theories of behavior change so that the programs can be designed to change thosernfa ctors which lead to the undes irable ri sk behaviors and maintain the des irable once. With thernobjectives of determining I-ITV/AIDS related sex ual behaviors and assessing their relation withrnvariables derived from the classical models of hea lth behavior, a study using a pre-tested se lfadministeredrnquestiOlUlaire to collect information from a random sample of six hundred and sixtyrnhigh school students was conducted in Jimma town on March 2000. Among the high schoolrnstudents 6.7% of female students and 11 ,2%. of male students were found to have involved inrnsexual activity with worst lifetime sex ual behavior index. Of the sexually active students the raternof unprotected sexual intercourse and intercourse with more than one sexual partner in the pastrnyear were St. 7% and 19.3%, respectivel y.rnThere is a signifi cant gender difference in HIV-related ri sk sexual act and belief-intention factors .rnHigher propOltions of males were involved in these high-risk sexual acts. Independent andrnsignifi cant predi ctors of intention toÂ· limit sexual partners were perceived barriers and normativernbelief for female students and Â· normati ve bel ief and perceived behavioral control for malernstudents. On the other hand, knowing an individual who has/had HJV/AIDS, motivation torncomply, normative belief, and perceived behavioral control for females ; while only perceivedrnsusceptibi lity and perceived behavioral control for males independently and sign ificantlyrnpredicted the intention to use condom in the next sexual intercourse. It was concluded that highrnschool students are at ri sk of HIV infection because of the high-ri sk behaviors they have. Patternsrnof gender difference are found and theory-driven and gender appropriate educationalrninterventions and researches are recorilillended.