Dissolution · Worked example
How to calculate the f2 similarity factor (with a worked example)
The similarity factor f2 compares two dissolution profiles — usually a test and a reference product — on a single 0–100 scale. Values of 50 or more are conventionally read as "similar". Below: the formula, the conditions that must hold before f2 is meaningful, and a complete calculation you can reproduce.
Open the f2 calculator Compare full profiles with bootstrap
The formula
f2 = 50 · log10 { [ 1 + (1/n) · Σ (Rt − Tt)² ]^(−0.5) · 100 }
Rt and Tt are the mean percentages dissolved for the reference and test product at time point t, and n is the number of time points used. Identical profiles give f2 = 100; an average difference of 10 percentage points at every time point gives f2 ≈ 50. The companion difference factor is
f1 = [ Σ |Rt − Tt| / Σ Rt ] · 100
with values of 0–15 conventionally read as similar (Moore & Flanner, 1996).
Check the conditions first
f2 is only interpretable when the data meet the conditions set out in the FDA (1997) dissolution guidance and the EMA (2010) bioequivalence guideline. In summary:
- Use at least 12 units (dosage units) of each product.
- Use at least three time points, excluding zero, with identical sampling times and conditions for both products.
- Include no more than one time point after 85 % dissolution for both products.
- Keep variability low: coefficient of variation ≤ 20 % at the earlier time points and ≤ 10 % at the later ones. If this fails, a model-independent multivariate or bootstrap approach is more appropriate.
- Under the EMA guideline, if both products release more than 85 % within 15 minutes, the profiles are considered similar without calculating f2.
Exactly how "85 %" is applied (> 85 or ≥ 85) varies; SCIOVANTA asks you to choose the convention explicitly rather than assuming one.
Worked example
Mean cumulative release (%) for a reference and a test tablet:
| Time (min) | Reference R | Test T | R − T | (R − T)² |
|---|---|---|---|---|
| 5 | 18 | 14 | 4 | 16 |
| 10 | 35 | 30 | 5 | 25 |
| 15 | 52 | 46 | 6 | 36 |
| 20 | 66 | 61 | 5 | 25 |
| 30 | 82 | 79 | 3 | 9 |
| 45 | 93 | 91 | 2 | 4 |
| Sum | 346 | 25 | 115 |
Only the 45-minute point exceeds 85 % for both products, so all six points are retained (n = 6).
- Mean squared difference: 115 / 6 = 19.17
- 1 + 19.17 = 20.17; its square root is 4.491
- 100 / 4.491 = 22.27
- log10(22.27) = 1.348; × 50 = f2 = 67.38
- f1 = 25 / 346 × 100 = 7.23
Both values fall in the conventional "similar" range. In a real submission you would also report the number of units, the CV at each time point and the conditions above — the calculator lists each of these checks beside the result.
Common mistakes
- Including several points above 85 %: plateau points shrink the differences and inflate f2.
- Calculating f2 from individual units instead of the mean profiles.
- Ignoring high variability; with CV above the limits, f2 alone is not appropriate.
- Reporting f2 without the time points, units and convention used.
For variable data, a bootstrap confidence interval for f2 gives a more honest picture. The profile comparison tool provides percentile and BCa intervals from independent-unit resampling.
Worked numbers on this page are reproduced exactly by the SCIOVANTA calculators linked above. Acceptance criteria differ between regulators and products; check the current guidance that applies to your submission.
References
- U.S. FDA (1997). Guidance for Industry: Dissolution Testing of Immediate Release Solid Oral Dosage Forms.
- European Medicines Agency (2010). Guideline on the Investigation of Bioequivalence (CPMP/EWP/QWP/1401/98 Rev. 1), Appendix I.
- Moore JW, Flanner HH (1996). Mathematical comparison of dissolution profiles. Pharmaceutical Technology 20(6):64–74.
- Costa P, Sousa Lobo JM (2001). Modeling and comparison of dissolution profiles. European Journal of Pharmaceutical Sciences 13(2):123–133.
Last updated 2026-11-17.