I thoroughly enjoyed reading this paper. The dataset is extensive and precisely fills a long-standing gap in terrestrial records. This work will be highly impactful for researchers investigating deep-time terrestrial carbon cycles, hydrocarbon source rocks, and the interactions between life and climate on land. Its potential applications are broad, and I anticipate it will be widely utilized across diverse research fields. The manuscript is also clearly written and highly accessible, which is a significant strength. Only a few minor revisions are needed, all of which are straightforward to implement.

First, the figures feel a bit rough and basic right now. They do the job, but for a journal like ESSD that aims for high visibility, they could look a lot nicer. A round of polishing (cleaner colors, sharper layout, better fonts and legends) would make the whole paper feel more professional and attractive.

Second, age control is always the biggest challenge in terrestrial sediments. The authors explain their interpolation method clearly, but many readers will worry that it could create extra uncertainty or put samples in the wrong time bins. I’d strongly suggest adding a simple column (or at least a confidence level) for age uncertainty in the actual database file. It would also be good to add one short, honest paragraph in the Usage instructions section that directly tells people about this limitation and how to handle it. That small addition will make the dataset much safer and more useful.

Third, the data availability part feels a little incomplete. The Geobiology Database link is mentioned, but actually going online to browse or filter the data isn’t very convenient yet. The Excel file is easy to download, but keeping it updated over time could be tricky. I strongly recommend giving at least one additional, user-friendly and regularly updatable link (maybe a simple web dashboard or a clearly versioned live repository) so people can access and explore the data more smoothly and the authors can add new entries efficiently.

Minor points:

• In the abstract (line 13), “exclusively” feels a bit too strong because quite a lot of data come from transitional facies and wood/coal samples: “primarily” would be safer and more accurate.
• Line 19: “source-rock assessment” is better without the hyphen — just “source rock assessment”.
• Please carefully check consistency of field/attribute names between the main text and Table 1 (e.g., line 85 SampleID vs Sample ID, line 87 Strata vs Stratigraphic Unit).
• Around line 94 the text uses “PalaeoLat”, but Table 1 shows “Paleolat” — please pick one spelling and use it everywhere.
• In the attached data table there is an obvious typo: “Yancahng FM.” should be “Yanchang FM.” — worth a quick scan of the whole table to catch any others.

Overall, this is a rigorous and highly significant contribution that the field greatly needs. Following these modest revisions, the paper will be outstanding. I am pleased to recommend a minor revision and look forward to the revised manuscript.

Large-scale terrestrial plats colonized landscape since the Devonian, while the first angiosperm occurred in the Cretaceous. This work compiles terrestrial 49016 TOC and 17571 organic-carbon isotope data from Devonian to Cretaceous. These data cover the evolution processes of land plants and its associated organic-carbon isotope. It will facilitate the future studies of long-term carbon cycles and terrestrial ecosystem evolution. I recommend it to publish in this journal after minor revision.

1. The authors mention a sparse to nearly absent data organic-carbon isotope data in Devonian and Carboniferous. However, such data are very important for the study on plant evolution on land. Please try to get more data.
2. What are the other terrestrial facies besides lacustrine, deltaic and river?

3. Adding a figure of organic-carbon isotopic compositions curve from Devonian to Cretaceous would be interesting.

The authors present a database of 66.587 published TOC and carbon isotope measurements spanning 419-66 Ma, including robust metadata fields. The database will surely be of great benefit across many geoscience fields, both for current and future work.

The article itself is a short description that supports the publication of the data set. The data set is significant; it is unique, useful, and complete (see my one comment for Section 3.2), and the data set itself appears to be of high quality based on my own general screening and the description within.

Line comments:

37- this is dependent on depositional setting.

76- Database.,

Fig 1. Facies is spelled wrong (Faceies). Data Sources is spelled wrong (Data Sourcies).

146, 157- I’d caution use of academic theses and conference proceedings, which are not peer-reviewed. I suggest a filter for peer-reviewed vs not peer-reviewed datasets.

Fig 3. Is this by publication? So 619 bubbles?

Section 3.2. There is a discussion on why the dataset cuts off at 419 Ma, but not as to why the database cuts off at 66 Ma.

214- Can you list somewhere what facies fall under “other”?