Objectives: To identify perceptions among people with spinal cord injury (SCI) of the priorities for brain-computer interface (BCI) applications and design features along with the time investment and risk acceptable to obtain a BCI.

Design: Survey.

Setting: Research registry participants surveyed via telephone and BCI usage study participants surveyed in person before BCI use.

Participants: Convenience sample of people with SCI (N=40), consisting of persons from the registry (n=30) and from the BCI study (n=10). Participants were classified as those with low function (n=24) and those with high function (n=16).

Interventions: Not applicable.

Main outcome measures: Descriptive statistics of functional independence, living situations and support structures, ratings of importance of different task and design features, and acceptable levels of performance, risk, and time investment.

Results: BCIs were of interest to 96% of the low-function group. Emergency communication was the top priority task (ranked in the top 2 by 43%). The most important design features were "functions the BCI provides" and "simplicity of BCI setup." Desired performance was 90% accuracy, with standby mode errors no more than once every 4 hours and speeds of more than 20 letters per minute. Dry electrodes were preferred over gel or implanted electrodes (P<.05). Median acceptable setup time was 10 to 20 minutes, satisfying 65% of participants.

Conclusions: People with low functional independence resulting from SCI have a strong interest in BCIs. Advances in speed and setup time will be required for BCIs to meet the desired performance. Creating BCI functions appropriate to the needs of those with SCI will be of ultimate importance for BCI acceptance with this population.