Background Vitamin B-12 deficiency is common, and the incidence increases with age. Most people with vitamin B-12 deficiency are treated in primary care with intramuscular (IM) vitamin B-12. Doctors may not be prescribing oral vitamin B-12 formulations because they may be unaware of this option or have concerns regarding its effectiveness. Objectives To assess the effects of oral vitamin B-12 versus intramuscular vitamin B-12 for vitamin B-12 deficiency. Search methods We searched CENTRAL, MEDLINE, Embase, and LILACS, as well as the WHO ICTRP and ClinicalTrials.gov. The latest search date was 17 July 2017. We applied no language restrictions. We also contacted authors of relevant trials to enquire about other published or unpublished studies and ongoing trials. Selection criteria Randomised controlled trials (RCTs) comparing the effect of oral versus IM vitamin B-12 for vitamin B-12 deficiency. Data collection and analysis We used standard methodological procedures expected by Cochrane. Our primary outcomes were serum vitamin B-12 levels, clinical signs and symptoms of vitamin B-12 deficiency, and adverse events. Secondary outcomes were health-related quality of life, acceptability to patients, haemoglobin and mean corpuscular volume, total homocysteine and serum methylmalonic acid levels, and socioeconomic effects. We used GRADE to assess the quality of the evidence for important outcomes. We did not perform meta-analyses due to the small number of included trials and substantial clinical heterogeneity. Main results Three RCTs met our inclusion criteria. The trials randomised 153 participants (74 participants to oral vitamin B-12 and 79 participants to IM vitamin B-12). Treatment duration and follow-up ranged between three and four months. The mean age of participants ranged from 38.6 to 72 years. The treatment frequency and daily dose of vitamin B-12 in the oral and IM groups varied among trials. Only one trial had low or unclear risk of bias across all domains and outcome measures. Two trials reported data for serum vitamin B-12 levels. The overall quality of evidence for this outcome was low due to serious imprecision (low number of trials and participants). In two trials employing 1000 mu g/day oral vitamin B-12, there was no clinically relevant difference in vitamin B-12 levels when compared with IM vitamin B12. One trial used 2000 mu g/day vitamin B-12 and demonstrated a mean difference of 680 pg/mL (95% confidence interval 392.7 to 967.3) in favour of oral vitamin B-12. Two trials reported data on adverse events (very low-quality evidence due to risk of performance bias, detection bias, and serious imprecision). One trial stated that no treatment-related adverse events were seen in both the oral and IM vitamin B-12 groups. One trial reported that 2 of 30 participants (6.7%) in the oral vitamin B-12 group left the trial early due to adverse events. Orally taken vitamin B-12 showed lower treatment-associated costs than IM vitamin B-12 in one trial (low-quality evidence due to serious imprecision). No trial reported on clinical signs and symptoms of vitamin B-12 deficiency, health related quality of life, or acceptability of the treatment scheme. Authors' conclusions Low quality evidence shows oral and IM vitamin B-12 having similar effects in terms of normalising serum vitamin B-12 levels, but oral treatment costs less. We found very low-quality evidence that oral vitamin B-12 appears as safe as IM vitamin B-12. Further trials should conduct better randomisation and blinding procedures, recruit more participants, and provide adequate reporting. Future trials should also measure important outcomes such as the clinical signs and symptoms of vitamin B-12 deficiency, health related -quality of life, socioeconomic effects, and report adverse events adequately, preferably in a primary care setting.