Excel Solver Settings: How to Set Up and Choose Options

How-to basics · Updated November 18, 2026

In Excel for Microsoft 365 on Windows, open Data > Solver, set an objective cell, choose the cells Solver may change, add constraints, and select a solving method such as Simplex LP, GRG Nonlinear, or Evolutionary. The right Solver settings depend on whether your model is linear, nonlinear, or requires whole-number decisions.

What Solver settings control

Solver is an Excel add-in that searches for values in selected cells that make a formula reach a goal while satisfying rules you specify. The cells it changes are decision variables; the formula being optimized is the objective; and constraints define what is allowed. A production plan might maximize profit while staying within labor and material limits. A staffing model might minimize cost while meeting coverage requirements.

Before configuring Solver, organize the worksheet so the decision cells, objective formula, and formulas for each constraint are easy to identify. Solver changes cell values, recalculates the workbook, and evaluates whether the resulting objective and constraints meet the model. It does not infer the business meaning of the worksheet: the model and its assumptions are your responsibility.

Enable the Solver add-in

In Excel for Microsoft 365 on Windows, choose File > Options > Add-ins. At the bottom, set Manage to Excel Add-ins and select Go. Check Solver Add-in, then select OK. After it loads, Solver appears on the Data tab, usually in the Analysis group. If it is already enabled, go straight to Data > Solver.

On Excel for Mac, use Tools > Excel Add-ins, select Solver Add-in, and confirm. The exact placement can vary slightly with the Mac release, but Solver is opened from the Data tab once enabled. Excel for the web does not provide the desktop Solver add-in; open the workbook in the desktop app to use it. Menu paths and settings below refer to Excel for Microsoft 365 on Windows unless stated otherwise.

Set the objective and decision cells

In the Solver Parameters dialog, Set Objective is the cell containing the formula to optimize. Choose Max to find the largest value, Min to find the smallest, or Value Of and enter a target when the formula should reach a particular number. The objective must be a worksheet cell, generally one whose value changes when the decision cells change.

By Changing Variable Cells identifies the cells Solver may adjust. Start them with sensible numeric values; Solver uses them as an initial point, particularly for nonlinear models. Avoid including labels or formula cells among the decision variables. For example, a profit total should be a formula, not a changing cell.

Here is a small product-mix model. Put units to make in B2:B3, profit per unit in C2:C3, labor per unit in D2:D3, and material per unit in E2:E3. Enter profit values of 30 and 20, labor values of 2 and 1, and material values of 1 and 3. The available limits are 100 labor units and 90 material units.

With decision quantities in B2:B3 and unit profits in C2:C3, put this formula in B5 for total profit: =SUMPRODUCT(B2:B3,C2:C3). With labor requirements in D2:D3 and decision quantities in B2:B3, put this formula in B6 for labor used: =SUMPRODUCT(B2:B3,D2:D3). With material requirements in E2:E3 and decision quantities in B2:B3, put this formula in B7 for material used: =SUMPRODUCT(B2:B3,E2:E3).

In Solver, set the objective to B5 and choose Max; set changing cells to B2:B3; then constrain B6 to be less than or equal to 100 and B7 to be less than or equal to 90. With continuous quantities, the intersection of the two resource limits is 42 units of the first product and 16 of the second. That produces profit of 42 × 30 + 16 × 20 = 1,580, using exactly 100 labor units and 90 material units. If products must be made in whole units, add an integer constraint to B2:B3.

Add constraints that match the real rules

Select Add in Solver Parameters to define a constraint. The left side is usually a cell or range containing a formula, and the right side may be a number, a cell reference, or another range. Common relationships are less than or equal to, greater than or equal to, and equal to. In the example, the resource-use cells are constrained to be no greater than their available amounts.

  • Integer: decisions must be whole numbers, such as units, people, or machines.
  • Bin: decisions are binary, either 0 or 1, useful for yes-or-no choices.
  • AllDifferent: selected integer values must differ from one another, where that rule fits the model.
  • Nonnegative decisions: check Make Unconstrained Variables Non-Negative in Options if negative values are not meaningful. Otherwise, add explicit lower bounds where appropriate.

Do not add a rule just because it seems plausible. For example, a whole-number constraint is unsuitable for a model whose decision represents a divisible quantity such as hours or kilograms. Conversely, leaving it out can produce an impractical fractional answer when the real decision is indivisible.

Choose the solving method

Make Select a Solving Method at the bottom of the dialog match the formulas in the model. Solver’s three main methods are intended for different problem types:

  • Simplex LP: use for linear models, where the objective and constraints are linear functions of the changing cells. Examples include product mix and transportation models. The Assume Linear Model option can help Solver treat the model as linear, but it does not make nonlinear formulas linear; check that the worksheet formulas genuinely meet the linear assumption.
  • GRG Nonlinear: use for smooth nonlinear models, where formulas change continuously but not in a straight-line relationship. It is commonly used for formulas involving products of decision variables or smooth functions. Its result can depend on the starting values and may be a local optimum rather than the best possible value across every feasible point.
  • Evolutionary: consider for models with discontinuities, IF-based decision logic, or other formulas that are not smooth. It searches candidate solutions and may take longer. Results can depend on search settings and should be checked rather than treated as proof of a global optimum.

Models with integer or binary decisions need discrete constraints. The solving method and model structure affect how Solver handles those decisions, and mixed-integer models can take longer than a continuous linear model. For a straightforward linear model, start with Simplex LP; choose another method only when the mathematics requires it.

Understand the Options settings

Select Options in the Solver Parameters dialog to adjust tolerances and search behavior. Available options can differ depending on the solving method. For a first run, default values are often a reasonable starting point; changing them without a specific reason can make the search slower or the accepted result less precise.

  • Max Time and Iterations limit the search time and number of calculation steps. Raising a limit may let a difficult model continue longer, but does not guarantee a better answer.
  • Precision controls how closely constraints must be satisfied for Solver to treat a solution as feasible. It is not the number of decimal places displayed in a cell. Use appropriate number formats separately, and consider units and scale when choosing tolerance.
  • Use Automatic Scaling can help when model values differ greatly in size, such as costs in the thousands alongside proportions below one. It affects numerical handling, not the business rules.
  • Show Iteration Results pauses during the search to show intermediate results. This can help inspect a run but requires interaction and is usually unnecessary for routine calculations.
  • Integer Optimality sets an acceptable optimality gap for integer problems. A looser tolerance may allow Solver to stop sooner with a solution that is not proven as close to the optimum as a tighter tolerance would require.
  • Convergence is relevant to nonlinear searches and affects when Solver considers improvements small enough to stop. It is not a substitute for checking whether the model has reached a useful solution.

Other options may include estimates, derivative choices, and search behavior for nonlinear methods. These are advanced controls. Change them when you understand the model’s shape or are following a reasoned troubleshooting process, rather than assuming that a more aggressive setting always improves the answer.

Read the result and verify it

After selecting Solve, Solver reports whether it found a solution. You can keep the Solver solution or restore the original values. Depending on the model and method, Solver may also offer reports such as Answer, Sensitivity, or Limits. Report availability varies; for example, sensitivity analysis is associated with suitable linear models and is generally not available for integer-constrained problems.

Check the resulting decision cells, objective, and every constraint in the worksheet. Recalculate resource usage independently, confirm that units and signs make sense, and test nearby alternatives for a small model. A Solver message is not a validation of input data, assumptions, or real-world suitability. If a result is unexpectedly extreme, inspect formulas and limits before accepting it.

Common Solver problems

  • Solver is missing: enable the Solver Add-in through the add-in settings, then check the Data tab.
  • No feasible solution: the constraints may conflict, limits may be too restrictive, or formulas may not represent the intended quantities. Temporarily inspect rules one at a time and verify the units.
  • Solver stops without a useful result: confirm that the objective formula depends on the changing cells, select the right method, and review time or iteration limits. For GRG Nonlinear, try reasonable alternative starting values because the search can settle at a local optimum.
  • Linear model not recognized or a strange answer appears: inspect formulas for nonlinear operations, such as multiplying two changing cells, and ensure the selected method matches the model. Do not mark a model linear merely to force a particular method.
  • Fractional quantities appear: add an integer or binary constraint only for decisions that truly must be whole numbers. Such constraints can increase solve time.
  • Displayed values do not appear to satisfy a rule: increase displayed decimal places to inspect the underlying values, then review constraint precision. Cell formatting can hide decimals without changing the stored result.

Frequently Asked Questions

Where are Solver settings in Excel?
Enable the Solver Add-in if needed, then choose Data > Solver on Windows. Select Options in the Solver Parameters dialog for time, precision, scaling, and other method-dependent settings.

Which Solver method should I use?
Use Simplex LP for a genuinely linear model, GRG Nonlinear for smooth nonlinear formulas, and Evolutionary for discontinuous or nonsmooth models. The right choice follows the model’s mathematics, not simply the type of answer you want.

How do I make Solver return whole numbers?
In Solver Parameters, add an integer constraint for the decision cells that must be whole numbers. Use a binary constraint when a decision can only be 0 or 1. Avoid applying integer restrictions to quantities that can legitimately be fractional.

Why does Solver say it cannot find a feasible solution?
At least one constraint may conflict with another, or the formulas and limits may not describe a possible combination. Check cell references, signs, units, starting values, and the worksheet calculations before relaxing a rule.

Can I use Solver in Excel for the web?
The desktop Solver add-in is not available in Excel for the web. Open the workbook in Excel for Windows or Mac and enable Solver there to configure and run the model.

When Solver is not the right tool

Solver is useful when a worksheet has a defined objective, adjustable decisions, and explicit constraints. It is not a general-purpose forecasting tool, and it will not choose sensible assumptions on your behalf. For very large optimization problems, specialized optimization software or a dedicated service may be more appropriate than the standard Excel add-in. For a small or medium model, a clearly laid-out worksheet and carefully chosen settings make results easier to understand, reproduce, and verify.

Related

Try it: the Date Difference Calculator: days, months, years and weekdays between two dates, with the DATEDIF formulas.

More in How-to basics

Sources: Function behaviour and menu paths: Microsoft Support documentation for Excel. Method & sources

Spotted an error in this guide? Let us know.