This article provides a practical tutorial on performing a meta-analysis using the R statistical software. It describes how to conduct a meta-analysis with the freely available R package 'meta', using a working example from the field of mental health. The article also covers sensitivity analyses for missing binary outcome data and potential selection bias using the 'metasens' package. It outlines the essential R commands needed to conduct and report analyses, including conducting fixed effect and random effects meta-analyses, subgroup analysis, producing forest and funnel plots, and testing and adjusting for funnel plot asymmetry. The article concludes that R is a powerful and flexible tool for conducting meta-analyses and provides directions to more advanced methods available in R. The example used in the article is based on a Cochrane review evaluating the effect of haloperidol in the treatment of symptoms of schizophrenia. The results show that haloperidol is significantly more effective than placebo, but the prediction interval suggests that placebo might be superior in future studies. The article also discusses the impact of missing outcome data and the assessment and adjustment for small-study effects, including methods such as the trim-and-fill method and limit meta-analysis. The results indicate that missing data do not significantly affect the conclusion that haloperidol is better than placebo. The article concludes that R is a valuable tool for conducting meta-analyses and provides a brief overview of the statistical methods available in R for meta-analysis.This article provides a practical tutorial on performing a meta-analysis using the R statistical software. It describes how to conduct a meta-analysis with the freely available R package 'meta', using a working example from the field of mental health. The article also covers sensitivity analyses for missing binary outcome data and potential selection bias using the 'metasens' package. It outlines the essential R commands needed to conduct and report analyses, including conducting fixed effect and random effects meta-analyses, subgroup analysis, producing forest and funnel plots, and testing and adjusting for funnel plot asymmetry. The article concludes that R is a powerful and flexible tool for conducting meta-analyses and provides directions to more advanced methods available in R. The example used in the article is based on a Cochrane review evaluating the effect of haloperidol in the treatment of symptoms of schizophrenia. The results show that haloperidol is significantly more effective than placebo, but the prediction interval suggests that placebo might be superior in future studies. The article also discusses the impact of missing outcome data and the assessment and adjustment for small-study effects, including methods such as the trim-and-fill method and limit meta-analysis. The results indicate that missing data do not significantly affect the conclusion that haloperidol is better than placebo. The article concludes that R is a valuable tool for conducting meta-analyses and provides a brief overview of the statistical methods available in R for meta-analysis.