Try:

term = "whateverterm";
var encodeForLike = term => term.Replace("[", "[[]").Replace("%", "[%]");

string term = "%" + encodeForLike(term) + "%";
var data = conn.Query(@"
   select top 25 
  Term as Label, 
  Type, 
  ID 
  from SearchTerms 
  WHERE Term like @term", 
  new { term });

There is nothing special about like operators, you never want your params inside string literals, they will not work, instead they will be interpreted as a string.

note

The hard-coded example in your second snippet is strongly discouraged, besides being a huge problem with sql injection, it can cause dapper to leak.

caveat

Any like match that is leading with a wildcard is not SARGable, which means it is slow and will require an index scan.

Yes it does. This simple solution has worked for me everytime:

db.Query<Remitente>("SELECT * 
                     FROM Remitentes 
                     WHERE Nombre LIKE @n", new { n = "%" + nombre + "%" })
                   .ToList();

Best way to use this to add concat function in query as it save in sql injecting as well, but concat function is only support above than sql 2012

string query = "SELECT * from country WHERE Name LIKE CONCAT( % ,@name, % );"
var results = connection.query<country>(query, new {name});

The answer from Sam wasn t working for me so after some testing I came up with using the SQLite CONCAT equivalent which seems to work:

string sql = "SELECT * FROM myTable WHERE Name LIKE  %  || @NAME ||  % ";
var data = IEnumerable data = conn.Query(sql, new { NAME = Name });

Just to digress on Sam s answer, here is how I created two helper methods to make searches a bit easier using the LIKE operator.

First, creating a method for generating a parameterized query, this method uses dynamic: , but creating a strongly typed generic method should be more desired in many cases where you want static typing instead of dynamic.

public static dynamic ParameterizedQuery(this IDbConnection connection, string sql, Dictionary<string, object> parametersDictionary)
{
    if (string.IsNullOrEmpty(sql))
    {
        return null;
    }
    string missingParameters = string.Empty;
    foreach (var item in parametersDictionary)
    {
        if (!sql.Contains(item.Key))
        {
            missingParameters += $"Missing parameter: {item.Key}";
        }
    }
    if (!string.IsNullOrEmpty(missingParameters))
    {
        throw new ArgumentException($"Parameterized query failed. {missingParameters}");
    }
    var parameters = new DynamicParameters(parametersDictionary);
    return connection.Query(sql, parameters);
}

Then adding a method to create a Like search term that will work with Dapper.

public static string Like(string searchTerm)
{
    if (string.IsNullOrEmpty(searchTerm))
    {
        return null;
    }
    Func<string, string> encodeForLike = searchTerm => searchTerm.Replace("[", "[[]").Replace("%", "[%]");
    return $"%{encodeForLike(searchTerm)}%";
}

Example usage:

var sql = $"select * from products where ProductName like @ProdName";
var herringsInNorthwindDb = connection.ParameterizedQuery(sql, new Dictionary<string, object> { { "@ProdName", Like("sild") } });

foreach (var herring in herringsInNorthwindDb)
{
    Console.WriteLine($"{herring.ProductName}");
}

And we get our sample data from Northwind DB:

I like this approach, since we get helper extension methods to do repetitive work.

My solution simple to this problem :

parameter.Add("@nomeCliente", dfNomeCliPesquisa.Text.ToUpper());

query = "SELECT * FROM cadastrocliente WHERE upper(nome) LIKE " + " %" + dfNomeCliPesquisa.Text.ToUpper() + "% ";