Accuracy Calculator Guide — % Recovery, %RSD & Recovered Amount

The Four Steps on the Calculator Page

1. Enter the standard injections, standard weight and dilution factor.
2. Enter each level's spiked sample data.
3. Click Calculate % Recovery.
4. Review the verdict, per-level statistics, recovery chart and math log.

Every Setting, Explained

The page shows its default acceptance as a mean recovery of 98–102% and a %RSD per level of ≤ 2.0%, with custom limits supported.

Each spike level is its own small table, laid out like a paper or spreadsheet calculation sheet. The page starts with three levels, 80, 100 and 120, with three replicate rows in each.

Under every level, an Average value row gives the mean recovered amount and mean % recovery, and a % RSD row gives the level's %RSD.

• Add Spike Level adds a new level and suggests 20 more than the previous level, for example 140 after 80, 100 and 120. You can type any value.
• Add Replicate adds a row to that level, with the level value already filled in.
• Remove Level and the × on a row are available only while more than one level, or more than one row in that level, remains. After the first calculation, removing either recalculates straight away.
• Every row must be complete. Each replicate needs both its area and its spike weight, and every standard injection row needs a value. An empty row blocks the calculation exactly like a half-filled one. The empty cells are outlined in red, and you either fill them or remove the row.
• A row with a zero spike weight is skipped quietly: its result stays as a dash and it is not counted.

HPLC, standard weight 20.0 mg, dilution factor 10, no API factor, 3 decimal places, range 98% – 102%, %RSD limit ≤ 2.0%. Five standard injections: 1351242, 1349870, 1352510, 1350630 and 1351980.

Average Standard Area = 1351246.400, with SD 1049.917 and %RSD 0.078%.

For the first replicate of level 80: (1082400 ÷ 1351246.400) × 20.0 × 10 = 160.208 mg, then (160.208 ÷ 160) × 100 = 100.130%. This is the replicate the math log traces in its Step 2.

Pooling all nine replicates gives an Overall Mean % Recovery of 100.120% and an Overall %RSD of 0.106%. The mean is inside 98% – 102% and every level's %RSD is below 2.0%, so the banner reads Accuracy Passed.

When one replicate is off

Change the third area at level 100 from 1354100 to 1280000. That replicate now recovers 94.727%, and level 100 falls to a mean of 98.287% with a %RSD of 3.137%. The overall mean, 99.511%, is still inside 98% – 102%, but the level's %RSD is above the 2.0% limit, so the banner turns red and reads Accuracy Failed, with the reason that one or more levels exceed the %RSD limit. A good overall mean did not rescue it.

• Check 1, the range: the overall mean % recovery is at or above the lower end and at or below the upper end of the range you chose.
• Check 2, the %RSD of each level: every level that has two or more valid replicates must have a %RSD at or below your limit. A level with a single valid replicate shows n/a and is not checked.
• Both must be met for Accuracy Passed. The text under the banner then states the overall mean and the range, and that every level's %RSD is within the limit.
• If either check is missed, the banner reads Accuracy Failed and the text gives the reason, which can be the mean being outside the range, a level %RSD above the limit, or both.
• The Overall %RSD and the standard injections' %RSD are shown for your information. Neither is part of the verdict.
• A custom range needs a minimum that is lower than its maximum. A custom %RSD limit must be above zero and up to 100. Otherwise a message appears instead of a result.

• It starts with five rows, and Add Standard Injection (or Add Absorbance Reading for UV-Vis) adds more. The × removes a row as long as more than one remains.
• A live line shows the count, the average, the SD and the %RSD as you type.
• The page notes "limit ≤ 2% per USP <621>" beside that %RSD. The calculator displays the number but does not use it in the pass or fail verdict.
• At least 2 valid injection values are needed to calculate, and an average standard of zero stops the calculation.

The maths is identical. Only the labels and one check change. With UV-Vis selected, every absorbance you enter, in the standard injections and in the sample cells, must be between 0.00 and 2.00 AU. A reading outside that range is outlined in red as you type. If you switch to UV-Vis while larger values are already entered, they are flagged at once with a message to dilute, adjust or correct them. Calculating with an out-of-range value shows a message and no result.

• Verdict banner with the explanation line.
• Four result cards: Overall Mean % Recovery, Overall % RSD, Avg. Standard Area (from the standard injections) and Valid Data Points (n), the total number of replicates used.
• Level Summary table: level, conc. (%), n, mean % recovery and %RSD. A level's %RSD is shown in red when it is above your limit.
• Recovery chart: the % recovery of every replicate as dots, coloured by level, plotted by replicate number, between two dashed red lines at the lower and upper ends of your range.
• Math log in 4 steps: Step 1 the average standard area, Step 2 the recovery of the first valid replicate worked through, Step 3 each level's mean and %RSD, and Step 4 the overall mean and %RSD from the pooled replicates.

When the results refresh

Nothing appears until you press Calculate % Recovery. After that, changing the standard weight, dilution factor, API factor, decimal places, recovery range, %RSD limit, test type or instrument recalculates immediately, and so does removing a row or a level. Typing different standard injections, areas, spike weights or level values updates only the cell you edited. Press Calculate % Recovery again so the verdict, cards and chart match your data.

Copy All Results copies a text summary: a heading with the test and instrument, the overall mean % recovery, the overall %RSD, the average standard response, the acceptance range and %RSD limit, the verdict (PASSED or FAILED), a line for each level with its mean, %RSD and n, and "Calculated via BestPharmaTools.com". The button shows "Copied!" for a moment.

Print Report opens an editable A4 sheet built from your raw data. It lists the standard injections with mean, SD and %RSD, the standard weight, dilution factor, the API factor row only when you entered one, the instrument and the test name. Below that is a table for each level with every replicate's area, spike weight, recovered amount and % recovery, plus the average and %RSD rows, and a verdict line that reads Accuracy Passed or Accuracy Failed, or "Not yet calculated" if you have not calculated. You type the product name and the rest of the header and signature details directly on the sheet.

For a logo, use an image of 1 MB or less and no larger than 512 × 512 px. A bigger file or image is rejected with a message.

Clear Data restores five empty standard injection rows and the three default levels (80, 100, 120) with three replicates each, empties the weight, dilution factor and API factor, sets the test back to Accuracy (Assay) and the instrument back to HPLC, and hides the results and chart. Your decimal places, recovery range and %RSD limit selections stay as they are.

• Instrument type chosen, and at least 2 standard injections filled, with no empty injection rows.
• Weight of standard and dilution factor above zero.
• Every replicate row has both a sample response and a spike weight.
• The recovery range and %RSD limit are the ones your method SOP specifies.
• For UV-Vis, every absorbance is between 0.00 and 2.00.
• Select Test set to the right heading, so the report reads correctly.

How do I calculate % recovery in an accuracy study with this calculator?

First work out the Recovered Amount (mg) for each replicate: (Sample Area ÷ Average Standard Area) × Standard Weight (mg) × Dilution Factor, multiplied by the API Factor if you entered one. Then % Recovery = (Recovered Amount ÷ Spiked Amount) × 100. The calculator does this for every replicate row, then shows the mean and %RSD for each level and an overall mean recovery pooled across all levels.

What is the formula for recovered amount?

Recovered Amount (mg) = (Sample Area ÷ Avg. Standard Area) × Std Weight (mg) × Dilution Factor × Factor. The Factor part appears only when you type a value in the optional API Factor box. For UV-Vis, the same formula is used with absorbance in place of area.

What mean recovery range can I choose, and what %RSD limits?

The mean recovery range drop-down offers 98% – 102% (labelled Tight / ICH Q2 typical), 95% – 105% (Common Assay Spec), 90% – 110% (Wide / Trace-Level Spec), or a custom range. The %RSD limit per level offers ≤ 2.0% (Standard Spec), ≤ 1.0% (Tight Spec), ≤ 3.0% (Trace-Level Spec), or a custom limit. Use the ones that your method SOP specifies.

What decides Accuracy Passed or Accuracy Failed?

Two checks, both of which must be met. The overall mean % recovery must be within your chosen range, with the end values included. And the %RSD of every level that has two or more replicates must be at or below your %RSD limit. The overall %RSD is displayed, but it is not part of the verdict.

What is the difference between the level %RSD and the overall %RSD?

The level %RSD is calculated from the replicates of one spike level only, and it is compared with your %RSD limit. The overall %RSD pools every valid replicate from every level into one set and is shown for information. The overall mean is also worked out from this pooled set, not by averaging the level means.

How many spike levels and replicates can I use?

The page starts with three levels (80, 100 and 120) of three replicates each. You can add more levels with Add Spike Level, which suggests the previous level plus 20 as a starting value that you can change. Add Replicate adds a row to one level. You can remove levels and replicates as long as at least one level and one replicate row per level remain, and a calculation needs at least 2 valid replicates in total.

Does the Conc. of Sample (%) value change the calculation?

No. It is a label for the level. Typing it in one row copies it to every row of that level. It is shown in the level summary, chart, copied text and print report, but the recovery is calculated only from the area, standard values, dilution factor and the spike weight.

Why does a replicate row show a dash instead of a result?

A row is skipped when the area or spike weight is not a valid number, or when the spike weight is zero. Its Recover (mg) and % Recovery stay as a dash and it is not counted in the level statistics or in n. Check the spike weight of that row.

What does the Select Test drop-down do?

It chooses a heading: Accuracy (Assay), Accuracy (Dissolution), Accuracy (Content Uniformity (CU)) or Accuracy (Related Substances / Impurity). It updates the headings on the page, the results and the printed report. The calculation and the verdict do not change.

Do I need to press Calculate again after I edit a value?

It depends on what you edit. After the first calculation, changing the standard weight, dilution factor, API factor, decimal places, recovery range, %RSD limit, test type or instrument type recalculates by itself, and so does removing a row or a level. Typing new standard injections, areas, spike weights or level values does not. Press Calculate % Recovery again after those edits.

What is the API Factor?

An optional correction factor, for example a potency or purity factor. Leave it blank to skip it. When a value is entered, it is multiplied into the Recovered Amount formula and shown in the formula and math log. If you type a value, it must be above zero.

Is my data stored or sent to a server?

The calculation, chart, copy text and print preview all run in your browser, and the calculator script does not send your entries to a server.

Educational use only. This guide explains how the Accuracy Calculator on this site works. It is not a validated system and does not replace your SOPs or your own review. Please read the disclaimer for how to use this site.