Investor behavior and investment activities are strongly influenced by the risk tolerance level of individual investors. International evidence suggests that lower risk tolerant investors are generally risk averse. However, their demographic characteristics and socioeconomic factors drive their risk attitudes. This study aims at investigating the critical role that age, gender, marital/social status, number of dependents, educational qualifications, employment and income status, savings pattern, future monetary planning, investments amount and returns from investments play in influencing risk tolerance and thereby finding whether the individual investors are risk averse or risk prone. To fulfill these objective 12 questions representing hypotheses were asked to 200 individual investors investing regularly in the Indian stock markets. A risk tolerance points scale is prepared to analyze the risk attitudes overall and each factor wise, and a Binary Logit Model is applied to validate these results. On an overall basis, this study finds that the responded investors have a lower risk tolerance level which makes them highly risk averse. In line with the hypotheses drawn, this study proves that aged investors are more risk averse than their younger, inexperienced counterparts; married investors with children and other dependents are more risk averse than their unmarried and with less dependents counterparts; higher education brings risk tolerance attitude and thereby makes investors risk prone; higher income and savings also decrease risk aversion whereas future planning approach increases risk aversion. It is also found under this study that higher investments amount and returns from such investments increase the risk tolerance level and thus reduces risk aversion of these investors. However, contradictorily with the undertaken hypotheses, this study finds that women investors are more risk prone than their male counterparts, and employment status of the respondents is immaterial in regard to their risk attitude. Binary Logit Model results also mostly validate the above results except that it finds no impact of number