In economics, a budget constraint represents all the combinations of goods and services that a consumer (or other decision-maker) can purchase given current prices and a given level of income or wealth. In consumer theory, the budget constraint and a preference map (or system of indifference curves) are the basic tools used to analyse consumer choice. In the standard two-good case, the budget constraint can be represented graphically as a straight line showing the trade-off between the two goods. If ${\displaystyle x}$ and ${\displaystyle y}$ denote the quantities of two goods, with prices ${\displaystyle P_{x}}$ and ${\displaystyle P_{y}}$, and ${\displaystyle m}$ denotes income, the budget line is given by

Solving for ${\displaystyle y}$ yields

where ${\displaystyle m/P_{y}}$ is the vertical intercept (the maximum amount of ${\displaystyle y}$ the consumer can buy if ${\displaystyle x=0}$) and ${\displaystyle -P_{x}/P_{y}}$ is the slope of the budget line, representing the opportunity cost of one more unit of ${\displaystyle x}$ in terms of ${\displaystyle y}$ forgone. Similar constraints appear in models of labour–leisure choice, intertemporal consumption, firm behaviour and international trade.

In microeconomic consumer theory, the budget constraint is combined with a description of preferences to study individual utility maximisation. For a given income and price vector, the budget set (all bundles on or below the budget line) represents all the consumption bundles an individual can afford.

A common assumption is that preferences are well behaved: they are complete, transitive and monotonic, so that "more is better", and that indifference curves are downward-sloping and convex. Under these conditions, the individual's most preferred affordable bundle typically lies at a point where an indifference curve is tangent to the budget line. This tangency point represents the quantities of the two goods that maximise utility subject to the budget constraint.

However, the optimal bundle need not always be an interior solution. If the tangency point implied by the first-order conditions lies outside the feasible set, the optimum will instead be a corner solution, at which the consumer consumes only one of the goods, as in the case of perfect substitutes. As income varies, the locus of interior tangency points is called the expansion path.

Introductory treatments typically assume a linear budget constraint as above. In more realistic settings, budget constraints may be kinked or non-linear because of taxes, subsidies, quantity discounts, rationing, welfare benefits or other institutional features; these cases can also be analysed with the same basic tools, but the geometry and optimality conditions may be more complex.

In labour economics, the same two-good framework is used to model the trade-off between leisure and consumption. A worker is assumed to have a fixed time endowment that can be allocated between hours of work and hours of leisure; labour income equals the wage rate times hours worked, plus any non-labour income. The wage can then be interpreted as the price of leisure: taking one more hour of leisure reduces labour income, and hence consumption possibilities, by one hour’s wage. The resulting labour–leisure budget line, together with preferences over consumption and leisure, underlies the standard derivation of an individual labour supply curve.