Past studies of individual Internet adoption and usage have been mostly empirical and in developed countries or in urban settings of developing countries. These have largely examined socio-economic factors such as age, earnings, education, in driving adoption and use. Several of the past studies were done at a time when the Internet was a novelty and was primarily considered as a source of information for enhancing knowledge. Over time, with developments in social media and ecommerce, Internet is considered an effective medium for social networking, enabling knowledge creation and exchange and enhancing economic benefits. Using theory of social capital and social cognition helps us to understand the drivers of Internet use from the perspective of outcome expectations and self-efficacy along the social, economic and knowledge dimensions. The primary research question is: What factors drive outcome expectations and self-efficacy in Internet use? Our study is based on a survey in two rural areas (Ranchi, Jharkhand, India) and (Guna, Madhaya Pradesh, India). We used theory to develop a survey instrument on Internet users for understanding the drivers of Internet use based on outcome expectations and self-efficacy. We used data from the Principal Component Analysis (PCA) done previously, to identify the latent constructs as measures of outcome expectations and self-efficacy (Jain, 2016). Using ANOVA, the current study identified the differential across Age, Occupation, Digital Literacy, Earning, and Education on dimensions uncovered by PCA and related the findings to the rural context in a developing country. The PCA revealed three dimensions that were labelled as ‘Empowerment’, ‘Enhanced Scope of Work’ and ‘Transaction Efficacy’. There are statistically significant differences across those who are at different levels of Digital Literacy and Earnings and for ‘Transactional Efficacy’, in the two groups identified by type of Occupation as ‘Business’ and ‘Others’. Along the other two dimensions of ‘Empowerment’ and ‘Enhanced Scope of Work’, there is no statistically significant difference across these two categories of Occupation. Further, there are no statistically significant differences across different categories of Age and Education. Our results indicate that while a basic level of education may determine whether a user adopts Internet, once the user starts using the Internet with a goal orientation in terms of outcome expectations and self-efficacy, ‘Education’ level does not matter. A similar logic applies to ‘Age’. Since digitally literate users tend to have positive outcome expectations from Internet use, they may benefit far more than those who are not Digitally Literate. Therefore, public policy must not only focus on increasing Internet availability specifically in rural areas, there must be programs for increasing digital literacy as well. Without such support programs, Internet use outcomes would exclude those who are not as digitally literate. Since Internet is increasingly becoming the vehicle for economic growth, such exclusions could slow inclusive growth. Those with higher incomes had possibly higher levels of negative disconfirmations with Internet use than those with lower incomes. A similar logic applies for the ‘Transactional Efficacy’ component in the ‘Occupation’ category. The study identifies the possible drivers for the disconfirmations.