Studies involving the 30 s chair-stand test (CS-30) have shown that subjects’ movements can vary during the test, and that these variations may follow several patterns. The present study aimed to define these different patterns and their respective incidences among a population of community-dwelling older adults in Japan. We also investigated, among the patterns identified, potential associations with physical and mental characteristics. The study population comprised 202 community-dwelling older adults. Subjects were classified into four groups based on how their CS-30 performance (defined through sit–stand–sit cycle count) changed over three successive 10 s segments: “steady-goers,” “fluctuators,” “decelerators,” and “accelerators.” Several other measures were also evaluated, including sit-up count, knee-extension strength, toe-grip strength, and Mini-Mental State Examination score. We found that steady-goers and decelerators comprised 70% of the sample. Fluctuators and steady-goers showed comparable physical function. Decelerators exhibited significant correlations between CS-30 score (total cycles) and tasks involving persistence and repetitive actions (p < 0.05). In addition, accelerators showed significantly stronger knee extension than steady-goers (p < 0.01). Differences in temporal patterns of CS-30 performance corresponded to differences in certain dimensions of physical and mental function. Our findings may be useful for planning and evaluating intervention programs aimed at long-term-care prevention among community-dwelling older adults.