- ADAPTIVEREG — Multivariate adaptive regression splines
- CATMOD — Categorical data modeling
- FMM — Finite mixture models
- GAM — Generalized additive models
- GENMOD — Generalized linear models
- GLIMMIX — Generalized linear mixed models
- GLM — General linear models
- GLMSELECT — Performs effect selection in the framework of general linear models
- LIFEREG — Parametric models for failure time data that can be uncensored, right censored, left censored, or interval censored
- LOESS — Nonparametric method for estimating regression surfaces
- LOGISTIC — Models with binary, ordinal, or nominal dependent variables
- MIXED — General linear models with fixed and random effects
- NLIN — Nonlinear regression models
- NLMIXED — Nonlinear mixed models
- ORTHOREG — General linear models by the method of least squares
- PHREG — Regression analysis of survival data based on the Cox proportional hazards model
- PLM — Postfitting statistical analyses
- PLS — Principal components regression
- PROBIT — Maximum likelihood estimates of regression parameters and the natural (or threshold) response rate for quantal response data from biological assays or other discrete event data
- QUANTLIFE — Quantile regression models for survival data
- QUANTREG — Quantile regression models
- QUANTSELECT — Effect selection for linear quantile regression models
- REG — Ordinary least squares regression
- ROBUSTREG — Linear regression models in the presence of outliers
- RSREG — Quadratic response surface regression models
- SURVEYLOGISTIC — Models with binary, ordinal, or nominal dependent variables for complex survey sample designs
- SURVEYPHREG — Regression analysis of survival data based on the Cox proportional hazards model for complex survey sample designs
- SURVEYREG — Linear regression analysis for complex survey sample designs
- TPSPLINE — Penalized least squares
- TRANSREG — Linear models with optimal nonlinear transformations of variables

The REG procedure provides the most general analysis capabilities for the linear regression model. However, many other procedures can fit linear regression models, and many procedures are specifically designed for more general regression problems, such as robust regression, generalized linear regression, nonlinear regression, nonparametric regression, regression modeling of survey data, regression modeling of survival data, and regression modeling of transformed variables.

Below are highlights of the capabilities of the SAS/STAT procedures that perform regression analysis:

- univariate or multivariate linear least squares regression
- nine model-selection techniques including backwards, forwards, stepwise, and those based on R-squared
- diagnostics
- hypothesis tests
- partial regression leverage plots
- outputs predicted values and residuals
- graphics device plots

- finite mixture models
- response surface regression with estimation of factor levels for optimum response and ridge analysis
- multiple nonlinear least squares regression
- derivative-free
- steepest-descent, Newton, modified Gauss-Newton, Marquardt and DUD methods

- linear models with optimal nonlinear transformation
- high-accuracy regression by orthogonal transformations for ill-conditioned data
- smoothing splines
- multivariate adaptive regression splines
- maximum likelihood estimates of regression parameters for logit and probit models
- n:m conditional logistic regression
- robust regression and loess regression
- partial least squares
- generalized additive models
- Cox proportional hazards model
- parametric models for failure time data
- linear and nonlinear mixed models
- principal components regression
- penalized least squares
- quantile regression
- linear, logistic, and Cox proportional hazards regression for complex survey sample designs