you can see this 

R is a free software environment for statistical computing and graphics. It compiles and runs on a wide variety of UNIX platforms, Windows and MacOS. To download R, please choose your preferred CRAN mirror.

http://www.r-project.org/

The base functions in R tend to deal with missing values as most statistical packages do: omit them. Dropping incomplete records has the undesired side effect of potentially introducing bias. There are, however, packages that do varying degrees of imputation to deal with missing values. Imputation done inappropriately can also introduce bias. For example, simple single-variable imputation can alter the covariance structure.

In the end, it is the degree of missingness that becomes defining. For example, if the same variable is missing data from many observations and it is not well correlated with other variables, one might decide that there is less risk by dropping the variable than by imputing it. On the other hand if it is highly correlated it probably is redundant anyway.

Assuming you want to impute the missing data, you might want to look at these R packages:

-mi

-mice

-AmeliaII

I personally found AmeliaII easier, but it also depends on how you want to impute the missing data.