Working memory (WM) training has seen a surge of popularity in recent years. There is evidence for near transfer of training, but findings regarding far transfer have been inconclusive. Furthermore, concerns have surfaced regarding the methodological design used in many WM training studies. This paper examines these concerns and presents an improved experimental design. This experiment then tests three hypotheses - (a) WM capacity will improve as a function of training, (b) WM training facilitates near and far transfer, and (c) changes in different cognitive abilities will be correlated. The cognitive abilities measured before and after training were cognitive control (Stroop and AX-CPT) and reading comprehension (Nelson Denny and a selection of the GRE passage comprehension). Cognitive control served as a near transfer measure and reading comprehension as a measure of far transfer. The results show that WM training is associated with an improvement in cognitive control but not in reading comprehension. This finding is consistent with recent meta-analyses and suggests that WM training can induce near transfer but not far transfer.