Short answer: In one small randomized trial with metformin as a third arm, berberine improved weight, hormones and cholesterol more over 3 months, and myo-inositol improved blood sugar and insulin markers more.1 Inositol has more trials, but most of the evidence is low quality.2 Berberine lowered testosterone in pooled trials, with no solid evidence of more live births.3
| Berberine | Inositol (mostly myo-inositol) | |
|---|---|---|
| Head-to-head trial | 500 mg twice a day for 3 months: larger changes in weight, hormones and blood fats1 | 1,000 mg twice a day for 3 months: larger changes in blood sugar and insulin1 |
| Pooled trials | 12 trials: lower testosterone and LH/FSH ratio than placebo; lower cholesterol and waist size than metformin3 | 13 meta-analyses: better hormones, insulin resistance, ovulation and live births than placebo; mostly no different from metformin2 |
| Live births | Similar to placebo or metformin, lower than letrozole3 | More than placebo (risk ratio 2.29), on low-quality evidence2 |
| Pregnancy | No more serious pregnancy events than placebo in pooled trials3 | 2 g twice a day during pregnancy did not cut diabetes, preeclampsia or preterm birth (25.0% vs 26.8%)4 |
| Stomach side effects | No more than placebo (3 trials)3 | Not summarized in the review abstracts checked |
| Evidence quality | No solid evidence on clinical outcomes3 | No high-quality evidence; most low or very low2 |
In a randomized trial in India, women with PCOS took berberine 500 mg twice a day, metformin 500 mg twice a day, or myo-inositol 1,000 mg twice a day for 3 months. All three groups improved. Berberine showed the larger changes in body measures, hormones and blood fats, and myo-inositol the larger changes in blood sugar and insulin. The abstract does not give group sizes.1
In a 2026 study of 148 women treated in routine care (not randomized), groups on myo-inositol, berberine, metformin or both drugs each improved some measures, but after adjusting for differences at the start, the main metabolic results no longer differed between groups.5
Berberine. A 2019 meta-analysis of 12 trials found berberine lowered testosterone and the LH/FSH ratio versus placebo and lowered cholesterol and waist size versus metformin, but found no solid evidence that it improves live birth or other clinical outcomes.3
Inositol. A 2026 umbrella review of 13 meta-analyses found inositol improved hormones, insulin resistance, ovulation and live births versus placebo, but mostly matched metformin, and none of the evidence was rated high quality.2 A network analysis of 22 trials found myo-inositol plus D-chiro-inositol restored cycles better than metformin.6 In a JAMA trial of 464 pregnant people with PCOS, myo-inositol did not prevent pregnancy complications.4
The one randomized comparison is small and points different ways for different measures. Inositol has more trials; berberine trials are fewer and mostly small. PCOS care also involves cycles, fertility and long-term risks that a doctor can help track. This is general information, not personal advice. More detail: the Berberine and Inositol (Myo-Inositol) guides.