COBOL Program Structure

A COBOL program is organized as a strict hierarchy: DIVISIONS contain SECTIONS, sections contain PARAGRAPHS, and paragraphs contain one or more SENTENCES, each terminated by a period. This is unlike block-scoped languages such as Java or Python where curly braces or indentation define scope; in COBOL, structure comes from named, ordered units that must appear in a fixed sequence: IDENTIFICATION DIVISION, ENVIRONMENT DIVISION, DATA DIVISION, and PROCEDURE DIVISION. Understanding this hierarchy is essential before writing any real program, because misplacing a paragraph in the wrong division, or forgetting a required period, is one of the most common sources of compile errors for beginners.

A SECTION is an optional grouping within a division, declared with a name followed by the word SECTION and a period, for example FILE-CONTROL SECTION within ENVIRONMENT DIVISION. A PARAGRAPH is a user-named block of executable or declarative code, such as 100-CALCULATE-TOTALS, and a SENTENCE is one or more STATEMENTS (like MOVE, ADD, or PERFORM) ending in a period. Modern structured COBOL relies heavily on the PERFORM verb to call paragraphs like subroutines, for example PERFORM 100-CALCULATE-TOTALS, which lets programmers build modular, readable control flow instead of relying on the older, discouraged GO TO statement.

Traditional COBOL source files follow a fixed-format layout inherited from 80-column punch cards: columns 1-6 are reserved for optional sequence numbers, column 7 is the indicator area (an asterisk marks a comment line, a hyphen continues a literal), columns 8-11 form Area A (where division, section, and paragraph names must start), and columns 12-72 form Area B (where most statements are written). Columns 73-80 were historically reserved for identification text and are ignored by the compiler. Modern compilers like GnuCOBOL still support this fixed format by default, though many now also support a more flexible free-format mode enabled with a compiler directive.