FDA’s June 2026 guide sets a current dataset-size boundary and submission pattern for study-data files that exceed it.
ReviewedEvidence1 sourceSectionWriting & Style
Quick answer
FDA’s June 2026 guide says a study dataset should normally be a single transport file, but datasets larger than 5 GB should be split into smaller files of no more than 5 GB while also submitting the larger non-split file and documenting the split as directed.
Key details
Core IssueFDA’s June 2026 guide sets a current dataset-size boundary and submission pattern for study-data files that exceed it.
Registerscholarly, professional, or research/report writing
Important caveats
Scope Boundary
This is current FDA technical-conformance guidance for electronic study-data submissions; recheck the current FDA guide, supported standards, Data Standards Catalog, and eCTD specifications before relying on version-sensitive requirements.
Further guidance
Content
Keep each dataset in a single transport file when within the limit. For a dataset over 5 GB, provide the required smaller split files (each no more than 5 GB) in the designated split location together with the non-split file and document the split in the relevant reviewer guide.
Purpose
Keep very large datasets processable for FDA review while preserving the complete unsplit dataset and clear reviewer documentation.
Sources and evidence
Sources are shown with the role they play in this guide. Historical or style-sensitive claims are kept within the evidence boundary described above.
FDA’s June 2026 conformance guide sets maximum descriptor lengths of 8 characters for variable names and 40 characters for variable descriptive labels and dataset labels in the specified study-data exchange context. These numeric limits are current FDA technical-conformance specifications and should be rechecked when the guide or supported exchange format changes.
FDA’s June 2026 Study Data Technical Conformance Guide gives specific conventions for variable descriptive labels and dataset labels, including avoiding certain unbalanced punctuation that can interfere with processing. These requirements are current FDA technical-conformance guidance for electronic study-data submissions and should be rechecked when FDA guidance, supported standards, or eCTD specifications change.
FDA’s June 2026 conformance guide requires variable and dataset names to follow the applicable standards-development-organization technical specifications and gives SAS naming-option guidance for submitted datasets. These requirements are current FDA technical-conformance guidance for electronic study-data submissions and should be rechecked when FDA guidance, supported standards, or eCTD specifications change.
FDA’s June 2026 conformance guide recommends sizing character-data columns to the maximum variable length used across study datasets, with a separate individual-dataset rule for SUPPQUAL datasets. This is current FDA technical-conformance guidance for study-data files; exact implementation must still follow the applicable supported data standard and submission format.
FDA’s June 2026 conformance guide addresses character encoding for study-data variable names, dataset names, labels, and values and warns that unsupported or inconsistent characters can create review or processing problems. These requirements are current FDA technical-conformance guidance for electronic study-data submissions and should be rechecked when FDA guidance, supported standards, or eCTD specifications change.
FDA’s June 2026 conformance guide specifies .xml for XML files used in the covered study-data submission context and states that define.xml should not be compressed.
FDA’s June 2026 conformance guide notes that there is no single standard or best practice for all adjudication data and directs sponsors to document where adjudication data reside and how adjudicated values can be distinguished from original values.
