This paper presents the Global Preference Survey (GPS), a globally representative dataset on risk and time preferences, positive and negative reciprocity, altruism, and trust. The survey collected data from 80,000 individuals across 76 countries, representing about 90% of the world's population and global income. The data exhibit substantial variation across countries, which is partly systematic, following distinct economic, institutional, and geographic patterns. Individual-level heterogeneity in preferences is even more pronounced and varies systematically with age, gender, and cognitive ability. The preference measures are predictive of a wide range of individual behaviors, including savings, schooling decisions, labor market choices, health choices, prosocial behaviors, and family structure. The paper also explores the cultural origins of preference variation using language structure data. Key findings include: 1. **Cross-Country Variation**: Preferences vary substantially across countries, with systematic patterns related to economic, geographic, and cultural factors. 2. **Individual-Level Heterogeneity**: Most individual-level variation in preferences is within countries, with significant correlations between preferences and individual characteristics such as age, gender, and cognitive ability. 3. **Predictive Power**: Preferences are predictive of various economic behaviors, including savings, education, self-employment, smoking, and prosocial behaviors. 4. **Cultural Origins**: Language structure is a significant factor in explaining global preference variation, with people speaking languages that do not require explicit coding of the future being more patient, positively reciprocal, trusting, and altruistic. The GPS data provide a comprehensive and systematic investigation of preference heterogeneity and its determinants, offering insights into the cultural and biological mechanisms shaping human preferences.This paper presents the Global Preference Survey (GPS), a globally representative dataset on risk and time preferences, positive and negative reciprocity, altruism, and trust. The survey collected data from 80,000 individuals across 76 countries, representing about 90% of the world's population and global income. The data exhibit substantial variation across countries, which is partly systematic, following distinct economic, institutional, and geographic patterns. Individual-level heterogeneity in preferences is even more pronounced and varies systematically with age, gender, and cognitive ability. The preference measures are predictive of a wide range of individual behaviors, including savings, schooling decisions, labor market choices, health choices, prosocial behaviors, and family structure. The paper also explores the cultural origins of preference variation using language structure data. Key findings include: 1. **Cross-Country Variation**: Preferences vary substantially across countries, with systematic patterns related to economic, geographic, and cultural factors. 2. **Individual-Level Heterogeneity**: Most individual-level variation in preferences is within countries, with significant correlations between preferences and individual characteristics such as age, gender, and cognitive ability. 3. **Predictive Power**: Preferences are predictive of various economic behaviors, including savings, education, self-employment, smoking, and prosocial behaviors. 4. **Cultural Origins**: Language structure is a significant factor in explaining global preference variation, with people speaking languages that do not require explicit coding of the future being more patient, positively reciprocal, trusting, and altruistic. The GPS data provide a comprehensive and systematic investigation of preference heterogeneity and its determinants, offering insights into the cultural and biological mechanisms shaping human preferences.