Objectives The heel-rise test is used to assess the strength and endurance of the plantar flexors in everyday clinical practice. However, several factors may affect outcomes, including sex, age, body mass index and activity level. The aims of this study were to revisit the reliability and normative values of this test, and establish normative equations accounting for several factors. Design Cross-sectional observational study with test–retest. Setting Community. Participants Volunteers (n = 566, age 20 to 81 years). Interventions Subjects performed single-legged heel rises to fatigue, standing on a 10° incline, once on each leg. A subset of subjects (n = 32) repeated the test 1 week later. Reliability was quantified using intraclass (ICC) correlation coefficients and Bland–Altman plots {mean difference [95% limits of agreement (LOA)]}, whereas the impact of sex, age, body mass index and activity level on the number of heel rises was determined using non-parametric regression models. Results The test showed excellent reliability (ICC = 1.0 right leg, 1.0 left leg), with mean between-day differences in the total number of heel-rise repetitions of 0.2 (95% LOA −6.2 to 6.5) and 0.1 (95% LOA −6.1 to 6.2) for right and left legs, respectively. Overall, males completed more repetitions than females (median 24 vs 21). However, older females (age >60 years) outperformed older males. According to the model, younger males with higher activity levels can complete the most heel rises. Conclusions The heel-rise test is highly reliable. The regression models herein can be employed by clinicians to evaluate the outcomes of heel-rise tests of individuals against a comparable normative population.