Covariate-adjusted response-adaptive (CARA) randomization procedures in clinical trials are a class of precision medicine procedures that adaptively change randomization probabilities based on the history of treatment assignments, responses, and covariate profiles. The main idea is to optimize allocation proportions so that more patients are assigned to the better treatment for their particular covariate profiles. Such procedures are complicated and are typically based on modeling treatments and covariates, with interactions, in a generalized linear model, thus requiring the clinical trial to be quite large. We describe these designs and also propose an enrichment strategy using CARA procedures.