This study examined Americans' perceptions of immigrants as threats and their implications on immigration policy views as well as immigrants' actual involvement in crime. Images of immigrant groups result from the perceived threats they pose to the crime rate, economy, political power, and nativism (Blumer 1958). I argued that these perceptions result in opposition to immigrants and support for stronger measures to exclude undocumented immigrants. Of special interest for this study was the "criminal immigrant" stereotype. Previous studies demonstrate that immigrants are not highly crime-involved even when they experience additional stressors during their adaptation processes. Yet, according to Agnew's (1992) general strain theory, immigrants may be prone to criminality due to additional strains they experience while adjusting to the new country. However, many immigrants, through transnational activities maintain ties with family and friends overseas, thereby making the immigration experience less stressful. I argued in this study that immigrants' underinvolvement in crime is partly due to their transnational ties, which may serve a protective role as social support and thus condition the effects of strains. To examine the implications for policy views of perceptions of immigrants and immigrants' actual crime involvement, the 2004 General Social Survey (GSS) and the Children of Immigrants Longitudinal Survey (CILS) were used. The hypotheses were tested by conducting univariate, bivariate, and multivariate analyses. Overall, perceived immigrant threat affects opposition to immigrants and support for stronger measures to exclude undocumented immigrants. Among the various groups examined, the levels of opposition to immigrants differ from that of support for stronger measures to exclude undocumented immigrants. In terms of immigrants and crime, immigrants were not disproportionately involved in crime, as is widely believed by the American public. Contrary to hypotheses, however, immigrants' strains were not significant predictors of crime, and transnational ties did not condition the effects of strains on crime. It is recommended that future research be designed using more comprehensive data set(s) that represent and reflect the growing immigration population in the United States. Particularly, research should include measurements of micro-level social dynamics specific to immigrants such as additional measures of transnational ties and resilience.