Used tool value retention by brand
Across 189 cordless power tool models measured on Aug 16, 2026, a used tool asks a median of 74% of what the same part number asks new on eBay. By brand, Makita is lowest at 58% across 9 models and DeWalt highest at 80% across 49 models. Every figure is the median of live asking prices, not sold prices, and each model behind it carries at least 8 verified used and 8 verified new listings.
By brand
Lowest share of new first — that is the brand where buying used saves the most.
| Brand | Models | Median % of new | Middle half |
|---|---|---|---|
| Makita | 9 | 58% | 47–64% |
| Milwaukee | 68 | 71% | 64–77% |
| Craftsman | 14 | 74% | 69–79% |
| Ridgid | 24 | 75% | 69–86% |
| Ryobi | 21 | 78% | 74–89% |
| DeWalt | 49 | 80% | 73–89% |
Brands with fewer than 8 qualifying models are not listed. A median over four models reads as authoritative as one over sixty and is not.
Bare tools only, like for like
The table above mixes bare tools with kits, and brands differ in how much of each they sell under a single part number. Restricting to part numbers that are unambiguously the bare tool removes that difference — and widens the gap rather than closing it.
| Brand (bare tools) | Models | Median % of new | Middle half |
|---|---|---|---|
| Milwaukee | 67 | 71% | 64–77% |
| DeWalt | 14 | 85% | 76–92% |
By tool type
Pooled across brands. The spread here is narrower than the brand spread, which is itself the finding: what a tool is matters less than where it sits in the market.
| Tool type | Models | Median % of new | Middle half |
|---|---|---|---|
| Drills | 33 | 68% | 63–80% |
| Impact Wrenches | 9 | 73% | 65–83% |
| Reciprocating Saws | 22 | 73% | 66–77% |
| Oscillating Tools | 12 | 73% | 68–78% |
| Impact Drivers | 33 | 74% | 69–79% |
| Hammer Drills | 19 | 76% | 64–87% |
| Sanders | 9 | 77% | 72–80% |
| Circular Saws | 15 | 78% | 69–91% |
| Nailers | 8 | 86% | 80–87% |
What this measures, and what it does not
This is a price gap, not a depreciation rate. Both halves of the ratio are asking prices on eBay today: the median used listing over the median new listing for the same part number. It answers “how much do I save buying this used right now”. It does not track one tool’s value over time, and it is not a percentage of the manufacturer’s list price.
A low percentage is not proof a brand loses value quickly. Three different things pull it down, and this data cannot separate them: the tool genuinely holding value poorly, a cheap and heavily-discounted new market on eBay raising the denominator, or a deep and competitive used market lowering the numerator. Milwaukee carries several times more used listings per part number than the other brands here, so the third applies to it far more than to anything else. Read Makita’s 58% as “the used market for these is competitive”, not as “these depreciate fastest”.
Asking prices, not sold prices. eBay stopped exposing completed sales to anonymous requests, so no honest free source of sold data exists. Some listings here are priced optimistically and will never sell, which is why every row shows the middle half of its range beside the median rather than a bare point estimate.
“New” means new on eBay. That pool includes open-box and grey-market stock, so it is a street price rather than list. It is the right denominator for the question a buyer is actually asking, and it is a lower one than MSRP would be — which means these percentages are higher than a percentage-of-MSRP would show. Full detail on the methodology page.
Common questions
Which power tool brands hold their value best?
How much cheaper is a used power tool than a new one?
Does Milwaukee hold its value better than DeWalt?
Which used tools save the most against buying new?
Is this based on sold prices or asking prices?
Citing this
These figures are free to quote with attribution to Used Tool Index. They are recomputed daily from live listings, so please cite the date you read them — Aug 16, 2026 for this version. Per-model figures are on each tool’s page, and the full dataset is available as a single plain-text table.