Ecosystem service (ES) trade-offs arise from human management choices that affect the type, magnitude, and mix of services provided by ecosystems. Trade-offs occur when the provision of one ES is reduced due to increased use of another. These can be explicit or unintentional, and are classified along three axes: spatial scale (local vs. distant), temporal scale (rapid vs. slow), and reversibility (likelihood of recovery). Across four Millennium Ecosystem Assessment (MA) scenarios and case studies, trade-offs show a preference for provisioning, regulating, or cultural services, with supporting services often being overlooked. Cultural ES are largely unquantified in scenario models, leading to incomplete assessments of their losses. Quantitative models primarily capture provisioning and regulating services, neglecting cultural and supporting services. Trade-offs in space are common, often linked to provisioning services, such as agricultural production affecting water quality. Spatial trade-offs can have long-term, irreversible effects. In time, short-term decisions often prioritize immediate ES provision, ignoring future consequences. Examples include dryland salinization in Australia, where short-term agricultural gains led to long-term soil degradation. Reversibility varies, with some trade-offs being irreversible, while others can be reversed with proper management. Trade-offs across ES involve more than one service being affected. For example, managing forests for timber may impact water quality or recreational value. The decline of Gyps vultures in India illustrates how the loss of a single species can cascade into broader ecosystem service losses. Trade-offs in MA scenarios show varying priorities: Global Orchestration emphasizes provisioning, Order from Strength favors current use over future potential, Adapting Mosaic involves local management with declining trade-offs, and TechnoGarden prioritizes technological solutions, often undervaluing cultural services. Key lessons include the need for policies that consider ES trade-offs across multiple spatial and temporal scales, incorporate lessons from past decisions, and monitor both short-term and long-term impacts. Effective management requires recognizing the complexity of ecosystem interactions and developing policies that minimize trade-off effects. The total pressure on ES worldwide is expected to increase, necessitating sustainable practices and adaptive management strategies to ensure long-term ecosystem and human well-being.Ecosystem service (ES) trade-offs arise from human management choices that affect the type, magnitude, and mix of services provided by ecosystems. Trade-offs occur when the provision of one ES is reduced due to increased use of another. These can be explicit or unintentional, and are classified along three axes: spatial scale (local vs. distant), temporal scale (rapid vs. slow), and reversibility (likelihood of recovery). Across four Millennium Ecosystem Assessment (MA) scenarios and case studies, trade-offs show a preference for provisioning, regulating, or cultural services, with supporting services often being overlooked. Cultural ES are largely unquantified in scenario models, leading to incomplete assessments of their losses. Quantitative models primarily capture provisioning and regulating services, neglecting cultural and supporting services. Trade-offs in space are common, often linked to provisioning services, such as agricultural production affecting water quality. Spatial trade-offs can have long-term, irreversible effects. In time, short-term decisions often prioritize immediate ES provision, ignoring future consequences. Examples include dryland salinization in Australia, where short-term agricultural gains led to long-term soil degradation. Reversibility varies, with some trade-offs being irreversible, while others can be reversed with proper management. Trade-offs across ES involve more than one service being affected. For example, managing forests for timber may impact water quality or recreational value. The decline of Gyps vultures in India illustrates how the loss of a single species can cascade into broader ecosystem service losses. Trade-offs in MA scenarios show varying priorities: Global Orchestration emphasizes provisioning, Order from Strength favors current use over future potential, Adapting Mosaic involves local management with declining trade-offs, and TechnoGarden prioritizes technological solutions, often undervaluing cultural services. Key lessons include the need for policies that consider ES trade-offs across multiple spatial and temporal scales, incorporate lessons from past decisions, and monitor both short-term and long-term impacts. Effective management requires recognizing the complexity of ecosystem interactions and developing policies that minimize trade-off effects. The total pressure on ES worldwide is expected to increase, necessitating sustainable practices and adaptive management strategies to ensure long-term ecosystem and human well-being.