A lasso ( or ), also referred to as a lariat, riata, or reata (all from Spanish la reata), is a loop of rope designed as a restraint to be thrown around a target and tightened when pulled. It is a well-known tool of the American cowboy. The word is also a verb; to lasso is to throw the loop of rope around something. Although the tool has several proper names, such terms are rarely employed by those who actually use it; nearly all cowboys simply call it a "rope," and the use of such "roping." To most cowboys, the use of other terms — especially "lasso" — identifies the user as a layman.

Lasso is an application server and server management interface used to develop internet applications anf is a general-purpose, high-level programming language. Originally a web datasource connection tool, for Filemaker and later included in Apple Computer's FileMaker 4.0 and Claris Homepage as CDML, it has since evolved into a complex language used to develop and serve large-scale internet applications and web pages.

In statistics and statistical and machine learning, lasso (least absolute shrinkage and selection operator) (also Lasso or LASSO) is a regression analysis method that performs both variable selection and regularization in order to enhance the prediction accuracy and interpretability of the statistical model it produces. It was introduced by Robert Tibshirani in 1996 based on Leo Breiman’s Nonnegative Garrote. lasso was originally formulated for least squares models and this simple case reveals a substantial amount about the behavior of the estimator, including its relationship to ridge regression and best subset selection and the connections between lasso coefficient estimates and so-called soft thresholding. It also reveals that (like standard linear regression) the coefficient estimates need not be unique if covariates are collinear.