Safer sexual behaviour is vital in curbing the HlV epidemic in many developing countries where the epidemic is primarily transmitted through heterosexual sex. Reliable and valid assessment of factors related to HIV risk behaviours are important in testing behavioural theories as well as informing effective AIDS prevention programmes. The current study tests the reliability and validity of a modified version of the Condom Outcome Expectancy Scale (COES) among a sample of HI'-negative sexual partners (n = 100) of HIV-infected individuals in Pune, India, Chang Mai,Thailand and Kampala, Uganda. Internal consistency reliability was measured using Cronbach's alpha coefficient, which had a value of 0.80 for the entire sample, with site-specific values of. 0.78 for India; 0.75 for Thailand; and 0.79 for Uganda. Test-retest reliability was conducted to test the scale's stability over time with 60% of the sample, resulting in a Cronbach's alpha coefficient of 0.70. The scale's structure was explored by analyzing response scores on the items using principal components analysis, which yielded a two-factor solution. The study indicates the utility of a modified version of the widely tested COES across international settings. Such research is necessary in understanding intervention targets across international settings.