STATS

It depends somewhat on where you've got that value - I'll assume from linear regression unless told otherwise! The R^2 statistic measures what's called the "explained variation". In a regression equation you set up an equation (such as weight = beta_1 x height + beta_2 x sex) that attempts to explain the observed variation in a variable (here weight) in terms of other measured variables (here height and sex). The value of R^2 measures what proportion of the variation in the outcome is explained by the variation in the explanatory variables (and therefore, how much of the variation in the outcome has to be "soaked up" by the random error term in the regression). A value of 0.005 means that 1/200th of the variation in the outcome is explained by the explanatory variables in the regression. Not very much!