Scroll to a given field and press Ctrl+F or Ctrl+Enter or click the Search button or use the Tools > Set Search Filter command.
Filters specified for more than one field at the same time are assumed to be merged by the AND logical operator and narrow down the search results.

• The displayed Search dialog box enables you to specify the filter and its type.
  
  • Regular expressions
  • Starts with, text patterns with the * and ? masks
  • Equal, Not Equal, Greater than, Smaller than
  • Between
  • AND, OR
  • IsEmpty (also after removing whitespaces)
  • Similar ("fuzzy" searches)
  • Flagged records (with user-defined flags)
  • Formula (including over 300 functions and references to fields in one record)
  To search for the currently selected field value, click the "=" button or press Ctrl + =.
• The default filter type for text fields is Regular Expression, and for numeric fields: Equal.
• After a given filter type is used for some field, it becomes its default type.
• The default initial filter type can be changed in the Settings > Options dialog box.
• Allow empty matches and Match case are two options used by the Regular Expression filter type only.
  Empty matches occur for Regular Expression patterns like \A\z or a?.
• Selecting the default Search All Now option causes instant re-searching of all fields after you click OK. Turning it off can be helpful and can prevent unnecessary re-searching, especially when entering relatively many filters (with the limit being the same as the field/column limit: 16K) for large data sets. After you enter them, you just use the Search All Now (Ctrl+F9) command.

There are four filter types available for numeric and text fields:

Regular Expressions (RegEx)

The filter string is a data pattern based on standard regular expression syntax. You can click the "?" button to display a list of commonly used expressions. Some examples of regular expressions:

Pattern	Matches
abc	substrings "abc"
.bc	three-character substrings consisting of any character followed by "bc"
\Aabc	field contents starting with "abc"
abc\z	field contents ending with "abc"
\Aabc.*123\z	field contents starting with "abc" and ending with "123", with any number of other characters in between
abc\d\z	substrings ending with "abc" and one digit
^a\d+	field contents containing a line starting with "a" and at least one digit
^a\d*	field contents containing a line starting with "a" and zero or more digits
[ab]+c	substrings "abc", "aabc", "abbabc", etc., but not "c"
[ab]*c	substrings "abc", "aabc", "abbabc", etc., and "c"
[^ab]+c	substrings ending with "c" and not containing "a" or "b"
\w\d{2,3}	substrings consisting of one letter followed by 2 or 3 digits
\Aab\d{2,3}c	field contents beginning with "ab", two or three digits, and "c"
abc|xyz	substrings "abc" or "xyz"
\A\z	empty fields

For more information, see PCRE regular expression syntax summary.

Plain Text (Starts With, Equal, Not Equal, ...)

The filter string represents a plain text string that is compared against the whole field contents. This filter category includes a few variants: Pattern, Equal, Not equal, Greater than, Less than, Between, And and Or.
For the Starts with, And and Or variants, the filter string can be a prefix of the compared strings and can contain the wildcard "?" and "*" characters. Any non-wildcard "?" and "*" characters must be prefixed with a tilde (~).
The remaining variants perform exact text string comparisons. For Between, And and Or, the entered filter must be a list of, respectively, two or more elements separated by spaces.

Filter string	Matches
abc	field contents beginning with "abc" ("Starts with")
?bc	field contents beginning with any character followed by "bc" ("Starts with")
*abc	field contents ending with any character followed by "abc" ("Starts with")
*	all non-empty field contents ("Starts with")
50.1 100.2	values equal to or between 50.1 and 100.2 ("Between", US regional numeric settings)
12/1/2012 12/31/2012	dates equal to or between the specified ones ("Between", US regional date settings)
ab cd *ef	field contents equal to "ab", "cd", or ending with "ef" ("Or")

Formulas

The filter string represents a formula expression that can make use of all the built-in functions, operators and references to other record fields. The formulas are similar to those used in spreadsheets, with a few exceptions:

• cell references are replaced by field names,
• the range ":" operator is not available,
• direct references to other tables via the "!" and "_" operators are not available.

When searching, such a formula is evaluated for each record, and a given record is considered to meet the searching criteria if the formula returns a non-zero value.

To learn more about building formulas, see: Entering formulas.

Flagged Records

Searching for records flagged/marked by a given flag, which is a unique combination of field font and background colors.

Long Text, Images/Files and Code fields always use Regular Expressions to filter records. When filtering Images/Files fields, the file names are searched.

Clicking the Paste button displays a list of recently entered search strings.

To remove all specified search filters, click the (None) search key on the toolbar or use the Tools > Reset Searching command.

Scroll to a field or select a range of fields which should be searched and use the Tools > Find Duplicates command.
To find duplicated (whole) records, select an entire row by clicking the row heading or pressing Shift+Space.

To find the complement of the current record set (that is, all the records that are not included in the current record set), use the Tools > Find Complement command.

Use the Edit > Find (F3) command and, in the displayed Search Toolbar, enter the desired substring. The Find Previous/Next commands simply scroll to the subsequent found table or form cell values. The Find All command performs full-text filtering of Text and Numeric fields (Long Text fields are not included).
The Search Toolbar also enables you to replace found substrings. If you want to perform such a find-and-replace action for one field only, select the column by clicking its header.