We observe that the randomness-complexity of an algorithm effects the time-complexity of implementing a version of it that utilizes a weak source of randomness (through a randomness-extractor). This provides an additional motivation for the study of the randomness complexity of randomized algorithms. We note that this motivation applies especially in the case that derandomization is prohibitingly costly.