How To Highlight Physical Books in Readwise

Readwise can turn physical-book highlighting into searchable, spaced-repetition notes—but doing it well depends on how the highlights are captured and tagged. The core workflow is to photograph a single page (not two tightly packed pages) so OCR can reliably extract the highlighted sentence, then save that passage as a Readwise highlight tied to the correct book. Once the text lands in Readwise, it becomes part of the app’s daily review system, which boosts recall and retention compared with manually flipping back through a physical copy.

The process starts with a practical problem: physical highlights are easy to mark but hard to review quickly. In the example, a hockey analytics book from 2020 had yellow highlighter notes, yet revisiting them required page-by-page searching—too slow to be useful. The workaround was earlier manual journaling with page numbers and thoughts, but the goal now is to migrate those insights into Readwise so they show up in reminders and benefit from space repetition.

In Readwise, highlights can be added either by typing the passage or by using the “photo” route. The photo route relies on OCR, so the transcript emphasizes a key capture rule: when photographing, include only one page per image when pages are close together. If two pages appear in the same photo, OCR may struggle and misread where the text belongs, even if it eventually guesses a page number. After taking a photo, the user selects the extracted sentence, saves it as a highlight, and optionally adds a note such as the page number or a brief interpretation.

Organizing notes inside Readwise also matters. The transcript demonstrates adding a title to structure the imported content: by using the tag “.h1” (and suggesting “h2.h” for lower-level headings), Readwise can treat that line as a header. In the browser view, the header then appears as a section label—effectively creating a table-of-contents-like structure for the book’s highlights.

There’s a tradeoff. OCR can be faster for long passages, but it can introduce typos and formatting issues (including spacing and posture-related alignment problems). Typing may be slower but can be more accurate. The takeaway is not a single “best” method; instead, it’s a repeatable approach—photo-based OCR for efficiency, manual typing for precision—so readers can test which workflow fits their reading habits and tolerance for cleanup.