Numerous studies have demonstrated the beneficial effects of anodal prefrontal transcranial direct current stimulation (tDCS) on working memory. However, a large variability exists in the applied tDCS parameters and working memory outcome measures. Using a meta-modeling approach, we investigated the relationship between tDCS electric fields in the left prefrontal cortex and improvements in working memory performance. Using this approach, a vector of outcome measures is correlated with the tDCS-related electric fields across several studies. These performance-electric field correlations (PEC) are calculated for each spatial location of the grey matter. Extracting 354 data points from 67 studies, we compared the spatial maps of tDCS effects on I) working memory accuracy and speed (regardless of working memory type and time of assessment), II) verbal and visuospatial working memory (regardless of performance measurement and time of assessment), and III) performance during and after stimulation (regardless of performance measurement and working memory type). We found that accuracy improves when anodal tDCS is applied to inferior frontal regions (Brodmann area 47) while working memory speed benefits from stimulation to dorsolateral and anterior prefrontal areas (Brodmann areas 9/10). Furthermore, the beneficial effects of left prefrontal tDCS are exclusive to verbal working memory, with no improvements in visuospatial working memory. We also observed region-specific effects only for task performance during, but not after, stimulation. The results of this study elucidate the causal involvement of prefrontal regions in working memory and can help guide tDCS placement for therapeutic application in disorders that involve working memory deficits.