Objective: Because of preliminary results from in vitro studies, hydroxychloroquine (HCQ) and chloroquine (CQ) have been proposed as possible treatments for COVID-19, but the clinical evidence is discordant. This study aims to evaluate the safety and efficacy of CQ and HCQ for the treatment of COVID-19. Methods: A systematic review with meta-analysis was performed. An electronic search was conducted in four databases for randomized controlled trials that compared HCQ or CQ with standard-of-care. A complementary search was performed. A quantitative synthesis of clinical outcomes was performed using the inverse variance method adjusting for a random-effects model. Results: In total, 16 studies were included. The meta-analysis found no significant difference between intervention and control groups in terms of mortality at the most extended follow-up (RR = 1.09, CI95% = 0.99-1.19, p-value = 0.08), patients with negative PCR results (RR = 0.99, CI95% = 0.89-1.10, p-value = 0.86), or serious adverse events (RR = 2.21, CI95% = 0.89-5.47, p-value = 0.09). HCQ was associated with an increased risk of adverse events (RR = 2.28, CI95% = 1.36-2.83, p-value < 0.01). The quality of evidence varied from very low to high. Conclusion: There is no evidence that HCQ reduces the risk of death or improves cure rates in patients with COVID-19, but it might be associated with an increased risk of adverse events