diff --git a/Dynamic/SimulatableModels/GainSchedModel.cs b/Dynamic/SimulatableModels/GainSchedModel.cs
index 97f692df..84d2a6ec 100644
--- a/Dynamic/SimulatableModels/GainSchedModel.cs
+++ b/Dynamic/SimulatableModels/GainSchedModel.cs
@@ -241,6 +241,7 @@ public void SetModelParameters(GainSchedParameters parameters)
                 else*/
             {
                 double x_otherInputs = modelParameters.Bias;
+                double gainSched = givenInputs[modelParameters.GainSchedParameterIndex];
                 //nb! input may include a disturbance!
                 if (givenInputs != null)
                 {
@@ -250,7 +251,7 @@ public void SetModelParameters(GainSchedParameters parameters)
                             continue;
                         if (i < GetModelInputIDs().Length)//model inputs
                         {
-                            x_otherInputs += CalculateLinearProcessGainTerm(i, givenInputs[i]);
+                            x_otherInputs += CalculateLinearProcessGainTerm(i, givenInputs[i], gainSched);
                         }
                         else // additive inputs
                         {
@@ -279,19 +280,50 @@ public void SetModelParameters(GainSchedParameters parameters)
         /// </summary>
         /// <param name="inputIndex">the index of the input</param>
         /// <param name="u">the value of the input</param>
+        /// <param name="u_GainSched">the value of scheduling input</param>
         /// <returns>contribution to the output y, excluding bias and curvature contributions</returns>
-        private double CalculateLinearProcessGainTerm(int inputIndex, double u)
+        private double CalculateLinearProcessGainTerm(int inputIndex, double u, double u_GainSched)
         {
             double processGainTerm = 0;
+            int gainSchedModelIdx = 0;
+            for (int idx = 0; idx < modelParameters.LinearGainThresholds.Length; idx++)
+            {
+                if (idx == 0)
+                {
+                    if (u_GainSched < modelParameters.LinearGainThresholds[idx])
+                    {
+                        gainSchedModelIdx = idx;
+                        break;// jump out of for-loop
+                    }
+                    else
+                    {
+                        gainSchedModelIdx = idx+1;
+                    }
+                }
+                else if (idx == modelParameters.LinearGainThresholds.Length - 1)
+                {
+                    if (u_GainSched > modelParameters.LinearGainThresholds[idx])
+                    {
+                        gainSchedModelIdx = idx;
+                    }
+                }
+                else
+                {
+                    if (u_GainSched > modelParameters.LinearGainThresholds[idx-1] &&
+                        u_GainSched < modelParameters.LinearGainThresholds[idx])
+                    {
+                        gainSchedModelIdx = idx;
+                    }
+                }
+            }
+ 
             if (modelParameters.U0 != null)
             {
-                //TODO: use correct linear gain
-                processGainTerm += modelParameters.LinearGains.First()[inputIndex] * (u- modelParameters.U0[inputIndex]);
+                processGainTerm += modelParameters.LinearGains.ElementAt(gainSchedModelIdx)[inputIndex] * (u- modelParameters.U0[inputIndex]);
             }
             else
             {
-                //TODO: use correct linear gain
-                processGainTerm += modelParameters.LinearGains.First()[inputIndex] * u;
+                processGainTerm += modelParameters.LinearGains.ElementAt(gainSchedModelIdx)[inputIndex] * u;
             }
             return processGainTerm;
         }
@@ -309,6 +341,8 @@ private double CalculateStaticStateWithoutAdditive(double[] inputs, double badVa
             double x_static = modelParameters.Bias;
 
             // inputs U may include a disturbance as the last entry
+            double gainSched = inputs[modelParameters.GainSchedParameterIndex];
+
             for (int curInput = 0; curInput < Math.Min(inputs.Length, GetLengthOfInputVector()); curInput++)
             {
                 if (curInput + 1 <= modelParameters.GetNumInputs())
@@ -326,7 +360,7 @@ private double CalculateStaticStateWithoutAdditive(double[] inputs, double badVa
                         }
                         lastGoodValuesOfInputs[curInput] = inputs[curInput];
                     }
-                    x_static += CalculateLinearProcessGainTerm(curInput, curUvalue);
+                    x_static += CalculateLinearProcessGainTerm(curInput, curUvalue, gainSched);
                 }
             }
             return x_static;
diff --git a/Dynamic/SimulatableModels/GainSchedParameters.cs b/Dynamic/SimulatableModels/GainSchedParameters.cs
index cac33a42..447ae9ab 100644
--- a/Dynamic/SimulatableModels/GainSchedParameters.cs
+++ b/Dynamic/SimulatableModels/GainSchedParameters.cs
@@ -27,7 +27,6 @@ public class GainSchedParameters : ModelParametersBaseClass
 
         public FittingSpecs FittingSpecs = new FittingSpecs();
 
-
         /// <summary>
         /// A time constant in seconds, the time a 1. order linear system requires to do 63% of a step response.
         /// Set to zero to turn off time constant in model.
@@ -35,12 +34,23 @@ public class GainSchedParameters : ModelParametersBaseClass
         public double[] TimeConstant_s { get; set; } =null;
 
 
+
+        /// <summary>
+        /// The index of the parameter among the ModelInputIdx
+        /// </summary>
+        public int GainSchedParameterIndex = 0;
+
+
+        /// <summary>
+        /// Thresholds for when to use which gains
+        /// </summary>
+        public double[] TimeConstantThresholds { get; set; } = null;
+
         /// <summary>
         /// The uncertinty of the time constant estimate
         /// </summary>
         public double[] TimeConstantUnc_s { get; set; } = null;
 
-
         /// <summary>
         /// The time delay in seconds.This number needs to be a multiple of the sampling rate.
         /// Set to zero to turn of time delay in model.
@@ -51,6 +61,11 @@ public class GainSchedParameters : ModelParametersBaseClass
         /// </summary>
         public List<double[]> LinearGains { get; set; } = null;
 
+        /// <summary>
+        /// Threshold for when to use different gains
+        /// </summary>
+        public double[] LinearGainThresholds { get; set; } = null;
+
         /// <summary>
         /// An array of 95%  uncertatinty in the linear gains  (u-u0))
         /// </summary>
diff --git a/TimeSeriesAnalysis.Tests/Tests/PlantSimulatorMISOTests.cs b/TimeSeriesAnalysis.Tests/Tests/PlantSimulatorMISOTests.cs
index a6debc32..4b6aa390 100644
--- a/TimeSeriesAnalysis.Tests/Tests/PlantSimulatorMISOTests.cs
+++ b/TimeSeriesAnalysis.Tests/Tests/PlantSimulatorMISOTests.cs
@@ -127,8 +127,10 @@ public void SetUp()
 
             gainSchedParameters1 = new GainSchedParameters
             {
-                TimeConstant_s = new double[] { 10 },
+                TimeConstant_s = new double[] { 10,20 },
+                TimeConstantThresholds = new double[] { 2},
                 LinearGains = new List<double[]> { new double[] { 5 }, new double[] { 10 } },
+                LinearGainThresholds = new double[] { 2 },
                 TimeDelay_s = 0,
                 Bias = 0
             };
@@ -223,18 +225,21 @@ public void GainSched_Single_RunsAndConverges()
             var isOk = plantSim.Simulate(inputData, out TimeSeriesDataSet simData);
             Assert.IsTrue(isOk);
             SISOTests.CommonAsserts(inputData, simData, plantSim);
-            double[] simY = simData.GetValues(processModel1.GetID(), SignalType.Output_Y);
+            double[] simY = simData.GetValues(gainSched1.GetID(), SignalType.Output_Y);
 
-            Assert.IsTrue(Math.Abs(simY[0] - (1 * 50 + 0.5 * 50 + 5)) < 0.01);
-            Assert.IsTrue(Math.Abs(simY.Last() - (1 * 55 + 0.5 * 45 + 5)) < 0.01);
+            Assert.IsTrue(Math.Abs(simY[N/3-2] - 5) < 0.2,"first step should have a gain of 5");
+            Assert.IsTrue(Math.Abs(simY.Last() - 30) < 0.2, "third step should have a gain of 10");
+            //  Assert.IsTrue(Math.Abs(simY.Last() - (1 * 55 + 0.5 * 45 + 5)) < 0.01);
 
-            /*Plot.FromList(new List<double[]> {
-                simData.GetValues(processModel1.GetID(),SignalType.Output_Y_sim),
-                simData.GetValues(processModel1.GetID(),SignalType.External_U,0),
-                simData.GetValues(processModel1.GetID(),SignalType.External_U,1)
+            //y[k] = a y[k-1] + b u[k]
+            Shared.EnablePlots();
+            Plot.FromList(new List<double[]> {
+                simData.GetValues(gainSched1.GetID(),SignalType.Output_Y),
+                inputData.GetValues(gainSched1.GetID(),SignalType.External_U,0),
             },
-                new List<string> { "y1=y_sim1", "y3=u1","y3=u2" },
-                timeBase_s, "UnitTest_SingleMISO");*/
+                new List<string> { "y1=y_sim1", "y3=u1" },
+                timeBase_s, "GainSched_Single");
+            Shared.DisablePlots();
         }
 
 
diff --git a/TimeSeriesAnalysis.Tests/Tests/PlantSimulatorSISOTests.cs b/TimeSeriesAnalysis.Tests/Tests/PlantSimulatorSISOTests.cs
index 2b76d659..d3fd24c4 100644
--- a/TimeSeriesAnalysis.Tests/Tests/PlantSimulatorSISOTests.cs
+++ b/TimeSeriesAnalysis.Tests/Tests/PlantSimulatorSISOTests.cs
@@ -236,7 +236,7 @@ public static void CommonAsserts(TimeSeriesDataSet inputData,TimeSeriesDataSet s
                 Assert.IsTrue(firstTwoValuesDiff < 0.01, "system should start up in steady-state");
                 Assert.IsTrue(lastTwoValuesDiff < 0.01, "system should end up in stedy-state");
             }
-            Assert.AreEqual(simData.GetLength(), simData.GetTimeStamps().Count(), "number of timestamps shoudl match number of data points in sim");
+            Assert.AreEqual(simData.GetLength(), simData.GetTimeStamps().Count(), "number of timestamps should match number of data points in sim");
             Assert.AreEqual(simData.GetTimeStamps().Last(), inputData.GetTimeStamps().Last(),"datasets should end at same timestamp");
 
         /*    foreach (var modelKeyValuePair in plant.GetModels())