Objectives To identify clinically relevant subtypes of apathy in older persons using latent class analysis (LCA) and investigate the distribution of different characteristics across these subtypes. Measurements: Cross‐sectional data of 122 older persons (mean age 84 years, 60% female) participating in the general population‐based PROactive Management Of Depression in the Elderly study, with apathy according to a score of 14 or higher on the Apathy Scale, were included for LCA. All 14 items of the Apathy Scale were used as indicator variables. Several characteristics were examined including sociodemographics, depressive, and anxiety symptoms; global cognitive function; quality of life indicators; hazardous alcohol intake (drinking ≥ 14 consumptions per week); and perceived chronic pain. Results Three distinct LCA classes were found classifying 17%, 7%, and 76% of the participants respectively. Individuals in the first class had a higher level of education and were less likely to live alone. Those in the second class had higher apathy and depression scores, lived more frequently alone and used more alcohol. Individuals in the third class showed a lower level of education and worse cognitive function. In multivariable multinomial analyses, only a lower educational level and higher scores on the Apathy Scale were significant predictors for class membership. Conclusion Differences between LCA‐derived classes were minimal, suggesting that in a general population‐based cohort the Apathy Scale measures a homogeneous construct. Copyright © 2016 John Wiley & Sons, Ltd.