diff --git a/modules/ml/doc/mldata.rst b/modules/ml/doc/mldata.rst
index 8c415b1e20..b09303701c 100644
--- a/modules/ml/doc/mldata.rst
+++ b/modules/ml/doc/mldata.rst
@@ -3,7 +3,7 @@ MLData
 
 .. highlight:: cpp
 
-For the machine learning algorithms, the data set is often stored in a file of the ``.csv``-like format. The file contains a table of predictor and response values where each row of the table corresponds to a sample. Missing values are supported. The UC Irvine Machine Learning Repository (http://archive.ics.uci.edu/ml/) provides many data sets stored in such a format to the machine learning community. The class ``MLData`` is implemented to easily load the data for training one of the OpenCV machine learning algorithms. For float values, only the  ``'.'`` separator is supported.
+For the machine learning algorithms, the data set is often stored in a file of the ``.csv``-like format. The file contains a table of predictor and response values where each row of the table corresponds to a sample. Missing values are supported. The UC Irvine Machine Learning Repository (http://archive.ics.uci.edu/ml/) provides many data sets stored in such a format to the machine learning community. The class ``MLData`` is implemented to easily load the data for training one of the OpenCV machine learning algorithms. For float values, only the  ``'.'`` separator is supported. The table can have a header and in such case the user have to set the number of the header lines to skip them duaring the file reading.
 
 CvMLData
 --------
@@ -182,6 +182,20 @@ Sets the variables types in the loaded data.
 
 In the string, a variable type is followed by a list of variables indices. For example: ``"ord[0-17],cat[18]"``, ``"ord[0,2,4,10-12], cat[1,3,5-9,13,14]"``, ``"cat"`` (all variables are categorical), ``"ord"`` (all variables are ordered). 
 
+CvMLData::get_header_lines_number
+---------------------------------
+Returns a number of the table header lines. 
+
+.. ocv:function:: int CvMLData::get_header_lines_number() const
+    
+CvMLData::set_header_lines_number
+---------------------------------
+Sets a number of the table header lines.
+
+.. ocv:function:: void CvMLData::set_header_lines_number( int n )
+
+By default it is supposed that the table does not have a header, i.e. it contains only the data.
+
 CvMLData::get_var_type
 ----------------------
 Returns type of the specified variable
diff --git a/modules/ml/include/opencv2/ml/ml.hpp b/modules/ml/include/opencv2/ml/ml.hpp
index 6358fa7be5..24e4a4bc9f 100644
--- a/modules/ml/include/opencv2/ml/ml.hpp
+++ b/modules/ml/include/opencv2/ml/ml.hpp
@@ -2040,6 +2040,9 @@ public:
     const CvMat* get_responses();
     const CvMat* get_missing() const;
 
+	void set_header_lines_number( int n );
+	int get_header_lines_number() const;
+
     void set_response_idx( int idx ); // old response become predictors, new response_idx = idx
                                       // if idx < 0 there will be no response
     int get_response_idx() const;
@@ -2091,6 +2094,8 @@ protected:
     CvMat* var_idx_out; // mat
     CvMat* var_types_out; // mat
 
+	int header_lines_number;
+
     int response_idx;
 
     int train_sample_count;
diff --git a/modules/ml/src/data.cpp b/modules/ml/src/data.cpp
index 2b44d384be..9bf4674c4a 100644
--- a/modules/ml/src/data.cpp
+++ b/modules/ml/src/data.cpp
@@ -71,6 +71,7 @@ CvMLData::CvMLData()
 {
     values = missing = var_types = var_idx_mask = response_out = var_idx_out = var_types_out = 0;
     train_sample_idx = test_sample_idx = 0;
+	header_lines_number = 0;
     sample_idx = 0;
     response_idx = -1;
 
@@ -117,6 +118,17 @@ void CvMLData::clear()
     train_sample_count = -1;
 }
 
+
+void CvMLData::set_header_lines_number( int idx )
+{
+	header_lines_number = std::max(0, idx);
+}
+
+int CvMLData::get_header_lines_number() const
+{
+	return header_lines_number;
+}
+
 static char *fgets_chomp(char *str, int n, FILE *stream)
 {
 	char *head = fgets(str, n, stream);
@@ -153,9 +165,15 @@ int CvMLData::read_csv(const char* filename)
     if( !file )
         return -1;
 
-    // read the first line and determine the number of variables
-    std::vector<char> _buf(M);
+	std::vector<char> _buf(M);
     char* buf = &_buf[0];
+    
+	// skip header lines
+	for( int i = 0; i < header_lines_number; i++ )
+		if( fgets( buf, M, file ) == 0 )
+			return -1;
+
+    // read the first data line and determine the number of variables
     if( !fgets_chomp( buf, M, file ))
     {
         fclose(file);