Daily 24 h rhythms in bone turnover have been demonstrated but whether these rhythms are intrinsically generated circadian rhythms is not known. We thus aimed to investigate this using the commonly used constant routine protocol where external factors such as meals, activity, sleep/wake and light/dark are kept constant. Serum procollagen type I N-terminal propeptide (sPINP) (marker of bone formation) and C-terminal telopeptide of type 1 collagen (sCTX) (marker of bone resorption), were measured in 2 hourly blood samples taken sequentially across 26 h in healthy individuals (n = 22, aged 19-33 years, 50% female). Concentration of sCTX showed a cosine rhythm in all males (acrophase (peak) time (mean ± SEM) 02:48 ± 14 h:min, amplitude 0.15 ± 0.02 ng/mL). All of the females had a statistically significant cosine + linear fit (acrophase 03:24 ± 20 h:min, amplitude 0.05 ± 0.01 ng/mL). There was no sex difference in acrophase, but females had a significantly smaller amplitude (P < 0.001). For sP1NP, only 4 males (36%) and 1 female showed statistically significant rhythms for either cosine, or cosine + linear models. Overall, sCTX, but not sPINP, exhibited a robust circadian rhythm in both males and females. This finding suggests that the circadian clock regulation of bone resorption by osteoclasts is robust, whereas circadian clock regulation of bone formation by osteoblasts is minimal.