Neighborhood level of disorder has recurrently been identified as a strong predictor of neighborhood level of crime rates and residential fear of crime. However, as scholars have emphasized, it begs to question whether neighborhood disorder have been measured in a reliable way and with adequate tools. The main aim of this study is to evaluate to what extent virtual systematic social observations (SSO) through Google Street View (GSV) may reliably audit neighborhood physical disorder in comparison to self-reported levels of neighborhood disorder. Further, the study also intends to chart whether virtual SSO through GSV is a valid instrument by testing a fundamental notion of the Broken Windows theory. The study consists of two sets of data, i.e. virtual SSO through GSV of 21 census neighborhoods in the city of Malmö (Sweden) and self-reported data of respondents living in the particular neighborhoods of interest. The correlations between the methodological diverse constructs of neighborhood disorder are subsequently examined, as well as the correlation between virtually observed physical disorder and victimization of property crimes. The results suggests that virtually observed neighborhood physical disorder through GSV is significantly correlated to self-reported neighborhood level of disorder as well as to victimization of property crimes. Virtual SSO through GSV thus appears to be a reliable and somewhat valid alternative towards auditing neighborhood level of disorder, comparable to data gathered through a community survey. Virtual observations through GSV do however struggle with several limitations.