Physics is an experimental science: its laws are not deduced from logic, they are found by measuring. Every time we record pairs of values in the laboratory, the task is threefold: choose a mathematical model that describes the data, work out the model’s parameters from the data, and compare model and data to judge whether the model is adequate. This chapter develops the tools for doing so — in particular for the most common non-linear models, exponentials and power laws — and reflects on what it really means to “confirm” or “falsify” a law.