The Bias-Variance Tradeoff describes the balance between a model’s ability to capture underlying patterns (low bias) and its sensitivity to noise (low variance). High bias can cause underfitting, whereas high variance can cause overfitting. Achieving the right balance is key to good generalization.