The overall research objective was to identify the effects of reward systems (rewards and a penalty) on risk-taking behavior and performance (quality and time) of construction workers from different cultures (American, Asian, and Latin American cultures). This research used the sociotechnical system as the underlying, guiding scientific framework. The research found that Americans and Latin Americans had higher risk-taking behavior than Asians (p<0.01). No difference in risk-taking behavior was found between Americans and Latin Americans (p<0.05). Although culture may influence individuals' risk-taking behavior, the results from this study showed that risk-taking behavior could be altered and suppressed by providing individuals with the proper safety training, education, and safety equipment. Customized safety training for people from different cultures would be useful because the culture elements that contribute to high risk-taking behavior could be addressed. The results also showed that the effects of reward systems on risk-taking behavior were not statistically significant (p>0.1). One possibility that no difference was found may be because the tasks used in this study did not contain enough possibility for participants to take more risk. The effects of reward systems on risk-taking behavior may have been reduced by the low possibility of risky behavior. It is suspected that if the tasks contained more opportunities for participants to take risk, differences in risk-taking behavior would have been significant. The researcher concluded that risk perception is situation-specific and has an influence on the individual's risk-taking behavior on that particular situation but cannot be used to predict risk-taking behavior. Also, general locus of control and general self-efficacy cannot be used to predict risk-taking behaviors. These findings are consistent with many studies that explore locus of control (Iversen & Rundmo, 2002; Rolison & Scherman 2002; Crisp & Barber, 1995), and many researchers that suggested self-efficacy is situation specific (Murdock et al., 2005; Martin et al., 1995; Perraud, 2000; Slanger & Rudestam, 1997). This study also found no relationship between risk-taking behavior and productivity, for both time and quality.