High-Low Method Calculator
Enter the total cost and activity level at your highest and lowest periods to separate fixed costs from variable costs using the high-low method. You also get a total cost projection for any future activity level and a step-by-step breakdown of the calculation. Results update as you type.
What is the high-low method?
The high-low method is a cost-accounting technique for separating the fixed and variable components of a mixed (semi-variable) cost using only two historical data points: the period with the highest activity level and the period with the lowest. It produces a simple linear cost equation in the form Total Cost = Fixed Cost + (Variable Cost Per Unit x Activity Level), which you can then use to estimate costs at any future output level. Because it requires minimal data and no statistical software, it is widely taught in introductory management accounting courses as a quick first approximation.
How to apply the high-low method: the formula step by step
Step 1 - Identify the periods with the highest and lowest activity levels from your historical data (not necessarily the highest and lowest costs). Step 2 - Subtract the low-period total cost from the high-period total cost to get the change in cost. Step 3 - Subtract the low-period activity level from the high-period activity level to get the change in activity. Step 4 - Divide the change in cost by the change in activity; that quotient is the variable cost per unit. Step 5 - Multiply variable cost per unit by the activity level of either the high or low period, then subtract from that period's total cost; the remainder is the fixed cost. Step 6 - Use the equation Total Cost = Fixed Cost + (Variable Cost Per Unit x Units) to project cost at any future activity level.
Worked example
A small manufacturer records total production costs of $54,000 for 18,000 units in its busiest quarter and $30,000 for 10,000 units in its quietest. Change in cost: $54,000 - $30,000 = $24,000. Change in units: 18,000 - 10,000 = 8,000. Variable cost per unit: $24,000 / 8,000 = $3.00. Fixed cost: $54,000 - ($3.00 x 18,000) = $54,000 - $54,000 = $0. Wait - try the defaults in this calculator, which use slightly different numbers to show a non-zero fixed cost. The point of the worked example is the process: isolate the variable rate from the slope, then back out the intercept (fixed cost) by plugging into either data point.
Limitations and when to use a different method
The high-low method has two well-known weaknesses. First, it uses only two data points, so if either the high or low period is an outlier (caused by a one-time event such as a plant shutdown, seasonal spike, or data-entry error), the resulting cost equation will be misleading. Second, it assumes a perfectly linear relationship between cost and activity, which is rarely true over a wide output range. For a more statistically robust estimate, use least-squares regression (also called ordinary least squares or OLS), which fits the best straight line through all historical data points rather than just the extremes. When the cost-activity relationship is visibly non-linear, a more advanced model may be needed. The high-low method remains useful as a quick sanity check or when historical data is scarce.
Cost structure benchmarks by fixed-cost share
| Fixed cost share | Classification | Implication |
|---|---|---|
| Below 30% | Variable-heavy | Costs fall quickly when volume drops; low break-even risk |
| 30% to 60% | Balanced | Moderate sensitivity to volume changes; typical for many industries |
| Above 60% | Fixed-heavy | High break-even volume; large operating leverage, higher risk at low output |
A rough guide for interpreting your fixed/variable cost split. Actual healthy ranges vary by industry.
Frequently asked questions
What does the high-low method calculate?
It splits a mixed cost into its two components: a variable cost per unit (the rate at which total cost rises with each additional unit of activity) and a total fixed cost (the baseline cost that persists even at zero activity). From those two numbers it builds a linear cost equation you can use to forecast total cost at any planned output level.
Do I use the highest cost or the highest activity level?
Always use the highest and lowest activity levels, not the highest and lowest costs. If the period with the highest cost is not the same as the period with the highest output, the formula will give incorrect results. Identify the data points by their activity volume (units, hours, calls, deliveries, etc.) first, then read off the costs for those same periods.
Can variable cost per unit be negative?
In the real world, no. A negative result almost always means your high and low data points are reversed (you entered the high cost as the low or vice versa), or the cost data contains an error. Double-check that the higher activity level corresponds to the higher total cost. If your actual data shows a negative slope, the cost driver you chose may not be the right one.
What activity measure should I use: units, hours, or something else?
Use whichever measure has the closest causal link to the cost you are analyzing. For direct labour costs, machine hours or labour hours work well. For delivery costs, number of deliveries or distance driven is usually better. For a factory, units produced is the natural choice. The method is valid for any cost driver as long as you use the same measure consistently for the high and low periods.
What is operating leverage and how does the high-low result relate to it?
Operating leverage measures how sensitive profit is to changes in sales volume. A business with a high fixed-cost share (fixed-heavy) has high operating leverage: small increases in revenue produce large gains in profit, but small revenue drops produce large losses. A variable-heavy business has lower leverage and more predictable margins at different volumes. Your fixed/variable split from the high-low method tells you where your cost structure falls on that spectrum.
How is the high-low method different from regression analysis?
The high-low method uses just two data points (the extreme observations), while regression analysis uses all available historical data points to fit the best possible straight line (minimising the sum of squared errors). Regression is statistically more reliable and less sensitive to outliers, but it requires more data and typically a spreadsheet or statistical tool. The high-low method trades accuracy for simplicity, making it ideal for quick estimates or preliminary analysis.