This study investigated the effects of resistance training (RT) performed with different frequencies, including a follow-up period, on cardiorespiratory fitness in healthy older individuals. Eighty-eight men and women (69 ± 3 years, 167 ± 9 cm and 78 ± 14 kg) were randomly placed into four groups: training one- (M1 = 11, W1 = 12), two- (M2 = 7, W2 = 14), or three- (M3 = 11, W3 = 13) times-per-week or a non-training control group (MCon = 11, WCon = 9). During months 1–3, all subjects trained two-times-per-week while during the subsequent 6 months, training frequency was set according to the group. Oxygen consumption (cycling economy: CE), gross efficiency (GE), blood lactate concentrations (La) and heart rate (HR) were evaluated during a submaximal cycle ergometer test. Hemoglobin (Hb), hematocrit (Hct), heart rate (HRrest) and body composition by DXA were also measured at rest. Maximal strength was measured by a 1-RM leg press test. Most improvements in CE, GE, La and HR occurred in all groups during months 1–3. No additional statistically significant improvements were observed during months 4–9, although effect sizes for the change in CE and GE at higher workloads indicated a dose-response pattern in men (CE at 75 W: M1 g = 0.13, M2 g = − 0.58, M3 g = − 0.89; 100 W: M1 g = 0.43, M2 g = − 0.59, M3 g = − 0.68) i.e. higher training frequency (two- and three-times-per-week versus one-time-per-week) led to greater improvements once the typical plateau in performance had occurred. Hb increased in W1 and W2, while no changes were observed in Hct or HRrest. 1-RM increased from months 1–3 in all intervention groups (except M2) and from month 4–9 only in M3 and in all women intervention groups. During follow-up, maximal strength was maintained but cycling economy returned to the baseline values in all training groups. These data indicate that RT led to significant improvements in cardiorespiratory fitness during the initial 3 months of training. This was partly explained by the RT protocol performed but further improvements may require higher training frequency. These changes are likely to be originated by the improved cardiorespiratory functions rather than neuromuscular adaptations evidenced by a lack of significant relationship during the intervention as well as the divergent results during follow-up.