A line score is a combination of your ASVAB subtest scores that one Service uses to decide which jobs you qualify for. The AFQT decides whether you can enlist at all; line scores decide which jobs are open once you do. Each branch writes its own formulas and reports them on its own scale: Army and Marine Corps composites center on 100, Navy composites are plain sums, and the Air Force reports percentiles.
Your score sheet starts with standard scores for each subtest, on a scale with a mean of 50 and a standard deviation of 10 in the 1997 national sample of 18 to 23 year olds. Everything else is built from those numbers.
The AFQT uses only four of them. Section 2.4 of the 1997 score scale report by Daniel Segall defines it as AR + MK + 2VE, converted to a percentile from 1 to 99 (see how the AFQT is calculated). The applicant FAQ on officialasvab.com separates the two: in enlistment testing you receive "an AFQT score and Service composite scores, used for military job assignment."
Service composites pull in the technical subtests the AFQT ignores: General Science, Electronics Information, Auto Information, Shop Information, Mechanical Comprehension and Assembling Objects. The Army's recruiting site calls them "line scores" and says they "determine job opportunities across different career categories."
Two inputs in the composite formulas are not subtests. Verbal Expression (VE) combines Word Knowledge and Paragraph Comprehension, and Auto and Shop (AS) combines Auto Information and Shop Information. Every Army composite uses VE, and nine of the ten use AS.
On the 1997 scale, VE is not WK + PC. Equation 2.3 weights the two so that PC carries 0.61934 of the weight of WK, then puts the result back on the mean-50 scale. Written against your standard scores, VE = 0.654918 × WK + 0.405617 × PC - 3.026712, rounded. The Verbal Expression guide walks through it.
Equation 2.4 weights Auto Information and Shop Information equally: AS = 0.539247 × (AI + SI) - 3.924703, rounded. The report notes that operational scoring uses the rounded, whole-number VE and AS.
The Army calculates 10 composites. Nine use non-integer weights designed to predict success in training, and Table 2.7 of the report gives those weights for the 1997 metric. To compute one, multiply each input by its weight, add the products, subtract the constant and round. The result has a mean of 100 and a standard deviation of 20 in the 1997 youth population, so a 120 is one standard deviation above average.
MM leans hardest on Auto and Shop, with a weight of 0.87. CL gives General Science no weight at all and rests mostly on AR, VE and MK. ST puts its largest weight on VE. Every composite still uses at least six inputs, so the short lists of "subjects" on the Army's own site describe emphasis, not the full formula.
GT is the one Army composite with integer weights. Table 5.1 defines it as AR + VE, and section 2.5.3 adds that the Army rescales that sum to the same mean-100 scale: GT = 1.074292 × (AR + VE) - 7.443781, rounded. For sums from 67 to 133 the rescaled GT stays within 2 points of the plain sum: 100 stays 100, and 113 becomes 114.
The GT score guide covers the jobs that use it. The line score calculator applies the same rescaling, so its Army GT is the figure the report describes.
The Navy does not rescale anything. Section 2.5.4 says it reports composites directly as the "sum of standardized subtest scores," so a Navy composite's size depends on how many tests it adds. With every subtest at 50, a two-test composite lands at 100, a three-test composite at 150 and a four-test composite at 200.
Navy.com says only that "some specific jobs also require minimum scores." The Navy sets rating minimums and changes them, so get the current composite and cut score for each rating from a recruiter.
The Marine Corps uses four composites. Each starts as a plain sum, then equation 2.6 rescales it to a mean of 100 and a standard deviation of 20, using the moments in Table 2.6. Marine GT adds Mechanical Comprehension, which neither the Army nor the Navy GT includes.
A footnote in the report says the Marine Corps did not yet use CL operationally in 2004; the report included it in anticipation of its use.
Marines.com sets the enlistment floor at 31, or 50 with a GED or nontraditional diploma. It says your scores help match you to an MOS but does not post MOS minimums.
The Air Force uses four unit-weighted composites known as MAGE. Section 2.5.1 names them Mechanical, Administrative, General and Electrical; airforce.com now calls the last two General Aptitude and Electronics.
Unlike the other Services, the Air Force reports each sum as a percentile from 1 to 99, like the AFQT. Appendix C of the report printed the 1997 conversion tables. The Air Force does not post its current conversion, so treat a sum you add up yourself as a rough index and rely on the percentile your recruiter reports.
The Space Force's career pages state requirements in the same terms. As published by the Space Force in October 2026, Space Systems Operator asks for an "electronics minimum score of 60." Job minimums change, so confirm the current figure with a recruiter.
The 1997 report defines composites only for the Army, Navy, Marine Corps and Air Force. Coast Guard Recruiting Command spells out a formula in its PY25 officer application guide: GT is the "sum of Arithmetic Reasoning (AR) and Verbal (VE) scores," with no rescaling. The same guide requires every part of a composite to come from tests taken the same day.
That guide is written for officer applicants. The Coast Guard jobs guide collects the enlisted A-school requirements the Coast Guard publishes; ask a recruiter which composite and minimum apply to your rating.
Take one applicant with these standard scores: GS 52, AR 56, WK 58, PC 54, MK 50, EI 45, AI 42, SI 48, MC 47 and AO 51.
VE = 0.654918 × 58 + 0.405617 × 54 - 3.026712 = 56.86, which rounds to 57. AS = 0.539247 × (42 + 48) - 3.924703 = 44.61, which rounds to 45. The AFQT composite is 56 + 50 + 2 × 57 = 220, a percentile of 70 in Table 2.5.
The Marine Corps GT for the same sheet is simpler. The sum is AR + MC + VE = 56 + 47 + 57 = 160. Then 0.755294 × 160 - 13.289535 = 107.56, reported as 108. The Army GT for this sheet is 114 (AR + VE = 113, rescaled), so one applicant on one test day holds two different GT scores.
The AFQT never sees GS, EI, AI, SI, MC or AO. Two applicants with identical math and verbal scores get the same AFQT and can still qualify for different jobs. In the table, applicant A is the example above. Applicant B matches A on AR, WK, PC and MK but scores higher on the technical tests: GS 55, EI 60, AI 61, SI 59, MC 60 and AO 55.
Start from the job. Get the composite name and minimum from a recruiter, then find that composite's formula above and list the inputs with the biggest weights. Those are the subtests to study.
Weights compound through VE and AS. One extra standard point on AR adds about 1.07 points to Army GT, while one extra point on WK adds about 0.70, because WK reaches GT only through VE. For Army MM, raising both AI and SI by 5 points lifts AS by about 5.4 and MM by about 4.7. If Marine GT is your target, Mechanical Comprehension counts as much as AR.
Line scores change only when your subtest scores do, which means a retest. The program office lists waits of 1 calendar month after your first test, 1 more after the first retest and 6 after any later one; the retake policy guide covers the details.