Objective To investigate the effect of second-generation antipsychotics on cognitive function in patients diagnosed with schizophrenia or schizoaffective disorder. Method Multiple-treatments meta-analysis model. Results On cognitive composite score, sertindole was superior to clozapine, effect size (ES) 0.87; 95% CI: 0.12–1.63, quetiapine, ES 0.75; 95% CI: 0.00–1.49, and first-generation antipsychotics (FGAs), ES 0.89; 95% CI: 0.14–1.64. Analyses on each cognitive domain showed clozapine, ES 0.37; 95% CI: 0.00–0.74, olanzapine, ES 0.31; 95%CI: 0.02–0.59, quetiapine, ES 0.34; 95% CI: 0.03–0.64, and FGAs, ES 0.51; 95% CI: 0.18–0.83 performing poorer on verbal working memory than ziprasidone, as well as FGAs performing poorer than risperidone, ES 0.31; 95% CI: 0.04–0.58. On executive function, sertindole performed better than clozapine, ES 0.82; 95% CI: 0.06–1.58, olanzapine, ES 0.81; 95% CI: 0.07–1.55, quetiapine, ES 0.76; 95% CI: 0.02–1.51, ziprasidone, ES 0.90; 95% CI: 0.14–1.67, and FGAs, ES 0.83; 95% CI: 0.08–1.58. On processing speed, FGAs performed poorer than sertindole, ES 0.97; 95% CI: 0.02–1.91, and quetiapine, ES 0.36; 95% CI: 0.01–0.72. On long-term verbal working memory, clozapine performed poorer than olanzapine, ES 0.41; 95% CI: 0.06–0.76. On verbal fluency, FGAs performed poorer than olanzapine, ES 0.26; 95% CI: 0.01–0.50, and clozapine, ES 0.44; 95% CI: 0.06–0.81. Lastly, FGAs, ES 0.41; 95% CI: 0.04–0.78, and clozapine, ES 0.44; 95% CI: 0.05–0.83, performed poorer on visuospatial skill compared to olanzapine. Conclusion The meta-analysis was able to detect some trends in the data analyzed, but did not show any drug having a uniform positive cognitive profile.