specific identification
Imagine a luxury car dealer with five cars on the lot, each with its own price tag and its own paperwork. When she sells the silver one she paid 60,000 for, she knows exactly which car left and exactly what it cost her — no guessing, no averaging. That literal, one-by-one tracking is specific identification.
Specific identification is the cost-flow method that drops all assumptions and simply matches each item sold to its own actual cost. Instead of pretending the first or last or average unit was sold, you record the true cost of the very unit that walked out the door. If the dealer sells the car that cost 60,000, cost of goods sold is exactly 60,000, and the remaining four cars stay in inventory at their own individual costs. It is the most accurate method when you can genuinely tell units apart.
It matters because it gives the truest matching of cost to revenue — which is why it is used (and often required) for high-value, unique, serial-numbered items: cars, jewelry, fine art, real estate, custom machinery. Its weakness is twofold: it is impractical for thousands of identical small items (you cannot meaningfully identify one screw from another), and where items ARE distinguishable it can open the door to profit manipulation, since a seller could deliberately 'choose' to sell the high-cost or low-cost unit to nudge the reported profit.
A jeweller holds three rings costing 1,200, 2,000, and 3,500. She sells the 2,000 ring. Under specific identification, cost of goods sold is exactly 2,000, and ending inventory is the other two rings at 1,200 + 3,500 = 4,700 — each carried at its own real cost, no averaging.
Each item carries its own actual cost — the truest match, but only feasible for distinguishable goods.
Specific identification is most exact but, for distinguishable goods, lets a seller pick which unit's cost to record — opening a door to manipulating profit by 'choosing' to sell a high- or low-cost item.