We investigated whether sponsor-imposed publication restrictions for ClinicalTrials.gov trials were reasonable, based on consistency with Good Publication Practice 2 (GPP2). ClinicalTrials.gov trial record data were electronically imported (7-October-2012) and screened for eligibility (phase 2-4, interventional, recruitment closed, results available, first received for registration after 10-November-2009, any sponsor type, investigators not sponsor employees). Two authors categorized restrictions information as consistent or not consistent with GPP2, resolving discrepancies by consensus. Of the eligible trials (388/484, n=81,768 participants), 80.7% (313/388) had restrictions disclosed and 92.5% (311/388) were industry-sponsored. Significantly more trials had restrictions that were consistent with GPP2 than not (74.1% [232/313], n=55,280 participants vs 25.9% [81/313], n=19,677 participants; P<0.001). Reasons for inconsistency were insufficient, unclear, or ambiguous information (48.1% 39/81), sponsor-required approval for publication (35.8%, 29/81), sponsor-required text changes (8.6%, 7/81), and outright bans (7.4%, 6/81). Follow-up of trials with insufficient information and a contact email (response rate, 46.9% [15/32]) revealed 2 additional bans. A total of 776 participants had consented to trials that had publication bans. Many, but not all, sponsor-imposed publication restrictions disclosed on ClinicalTrials.gov may be considered reasonable. Sponsors should ensure restrictions are appropriately disclosed. Volunteers should be alerted to any restrictions before consenting to participate in a clinical trial.