Measuring Internet performance for home users can provide useful information for improving network performance. Such measurements typically require users to install special software on their machines, a major impediment to use. To overcome this impediment, we designed and implemented several scripting techniques to predict Internet performance within the tightly constrained sandbox environment of a Web browser. Our techniques are integrated into a Web site project called "How's My Network" that provides performance predictions for common Internet activities, with this thesis concentrating on the performance of online news, social networks, and online shopping. We started our approach by characterizing news sites to understand their structures. After that, we designed models to predict the user's performance for reading news online. We then implement these models using Javascript and evaluate their results. We find out that news sites share common characteristics in their structures with outliers for some. Predicting the page load time according to number objects coming from dominant domain, the one providing the most number of objects, gives more accurate predictions than using total number of objects across all domains. The contributions of this work include the design of new approaches for predicting Web browser performance, and the implementation and evaluation of the effectiveness of our approach to predict Web browser performance.