Over the past 15 years, bilateral investment treaties (BITs) have given rise to a surge of investment claims. Investment scholars tend to believe that this rise of claims has significantly affected the law-making behavior of states in newly concluded BITs. While some commentators assert that states have lowered the level of investment protection in response to claims, others argue that states have included new exceptions as well as protective obligations in their treaties to strike a better balance between competing values underlying BITs. This thesis places this debate on firmer empirical grounds. Using quantitative and qualitative methods, it investigates the impact of investment arbitration on substantive BIT rule-making. It finds that BITs have evolved over time to include more obligations as well as exceptions. States thus do not lower investment protection in more recent treaties. At the same time, both camps of investment law scholars tend to overestimate the effect of investment arbitration on rule-making as most newly included provisions in BITs pre-date the rise of investment arbitration. Indeed, a quantitative assessment of the impact of investment claims fails to show an effect on rule-making. A case study of Canada, Germany and Japan further suggests that countries differed starkly in their reaction to investment claims. While Canada innovated its treaty-network and Germany continued business as usual after being hit by investment claims, Japan changed its treaty design without ever being a respondent in investment disputes. The best empirical evidence that investment arbitration impacts rule-making consists of specific clarifications that states have inserted into their investment treaties in response to investment claims. The thesis thus shows that investment arbitration has only led to a partial adjustment of BITs rather than a legal revolution. It thereby motivates the need to investigate other causes of change in investment treaty-making in future studies.