In regression analysis, collinearity of two variables means that strong correlation exists between them, making it difficult or impossible to estimate their individual regression coefficients reliably.
The extreme case of collinearity, where the variables are perfectly correlated, is called singularity.