OBJECTIVE A recent meta-analysis identified processing speed inefficiency as the largest single cognitive impairment in schizophrenia. However, the effect of potential moderator variables, such as medication status and severity of illness, remained unclear. The authors present an extended meta-analysis of processing speed and other specific cognitive functions in schizophrenia and examine the role of potential moderator variables. METHODS In addition to the studies identified in the original analysis, subsequently published articles were identified via systematic searches of MEDLINE and PsycINFO for the period from May 2006 to January 2009. The authors extracted data for potential moderator variables, including publication year; severity of illness; chlorpromazine equivalent daily dose; and mean IQ, mean age, and sample size for each study. Effect sizes were calculated for all measures, and meta-influence and homogeneity analyses were conducted. RESULTS Eleven studies were added to the original analysis, increasing the schizophrenia sample size from 1,915 to 4,135. The largest effect size was for coding tasks (g=-1.50), followed by category fluency (g=-1.31). However, for coding tasks, variation in effect size magnitude attributable to heterogeneity was substantial. Metaregression analyses indicated that three moderator variables were related to coding task effect size: publication year, IQ difference from comparison subjects, and chlorpromazine equivalent daily dose. There was a difference of 0.8 effect size units between studies with low compared with high chlorpromazine equivalent daily dose. No significant relationships were found between any moderators and the other cognitive tasks included in the meta-analysis. CONCLUSIONS The processing speed impairment in schizophrenia is substantially affected by several moderating factors, in particular antipsychotic medication dosage.