Abstract Background Right ventricular pacing (RVP) may cause electrical and mechanical desynchrony leading to impaired left ventricular ejection fraction (LVEF). We investigated the outcomes of RVP with His bundle pacing (HBP) and left bundle branch pacing (LBBP) for patients requiring a de novo permanent pacemaker (PPM) for bradyarrhythmia. Methods and results Systematic review of randomized clinical trials and observational studies comparing HBP or LBP with RVP for de novo PPM implantation between 01 January 2013 and 17 November 2020 was performed. Random and fixed effects meta-analyses of the effect of pacing technology on outcomes were performed. Study outcomes included all-cause mortality, heart failure hospitalization (HFH), LVEF, QRS duration, lead revision, atrial fibrillation, procedure parameters, and pacing metrics. Overall, 9 studies were included (6 observational, 3 randomised). HBP compared with RVP was associated with decreased HFH (risk ratio [RR] 0.68, 95% confidence interval [CI] 0.49–0.94), preservation of LVEF (mean difference [MD] 0.81, 95% CI − 1.23 to 2.85 vs. − 5.72, 95% CI − 7.64 to -3.79), increased procedure duration (MD 15.17 min, 95% CI 11.30–19.04), and increased lead revisions (RR 5.83, 95% CI 2.17–15.70, p = 0.0005). LBBP compared with RVP was associated with shorter paced QRS durations (MD 5.6 ms, 95% CI − 6.4 to 17.6) vs. (51.0 ms, 95% CI 39.2–62.9) and increased procedure duration (MD 37.78 min, 95% CI 20.04–55.51). Conclusion Of the limited studies published, this meta-analysis found that HBP and LBBP were superior to RVP in maintaining physiological ventricular activation as an initial pacing strategy.