Background/purpose To identify the unique predictors of comfortable and fast gait speed in community-dwelling older adults using measures of physical performance (eg, lower extremity strength and balance), self-reported balance confidence, and global cognitive function. Methods Demographic information was collected from 60 healthy, community-dwelling older adults older than 60 years. Participants completed the following assessments: Mini-Mental State Examination; Activities-Specific Balance Confidence Scale; 30-second Chair Stand (30-SCS); Functional Reach (FR); and gait speed (comfortable and fast) using the GAITRite system. Hierarchical linear regression was used to examine the relationship of both fast and comfortable gait speeds with functional performance (CST and FR), cognition (Mini-Mental State Examination), Activities-Specific Balance Confidence Scale, and demographic information (age, gender, and body mass index). Results and discussion Functional performance measures (30-SCS and FR) explained 55.4% and 64.7%, respectively, of the variance in comfortable and fast gait speed. Unique predictors for comfortable gait speed included 30-SCS, FR, and body mass index. Unique predictors of fast gait speed included 30-SCS, FR, gender, body mass index, and Activities-Specific Balance Confidence Scale score. These predictors explained 68.5% and 80.4% of the total variance in comfortable and fast gait speed, respectively. Global cognition was not a unique predictor of gait speed when performance measures were statistically controlled. However, the current study measured global cognitive status rather than the specific cognitive constructs, such as processing speed or executive function. Further research is needed to determine the role of cognition in the composition of gait speed. Conclusions Gait speed is an important indicator for many outcomes such as fall risk, mortality, and functional status. Understanding that key variables of strength and balance comprise a large portion of gait speed allows clinicians to better direct their care and optimize rehabilitation outcomes. This study specifically used functional measures of strength and balance that can be easily implemented in the clinical setting.