Circadian rate of intake and rumen parameters were studied in three rumen-fistulated steers fed long timothy hay once daily at 13:00 h. A sinusoidal model was used to relate rate of intake (INTAKE, kg h−1) (Y) to the rate of passage of small particles from the rumen. Parameters of the model Y = χ0 + χcos (time + θ) were the fitted mean intake at zero time (χ0), the fluctuation of intake (χ) and the phase. INTAKE was recorded seven times daily for the duration of the experiment. At the end of each week, fluctuations in individual INTAKE were regressed against recording hour over 24 h. The 7-wk experiment was divided into three periods. In the first period, fecal particles were examined. Small particles (< 0.8-mm diameter) represented 86 ± 4% of fecal NDF with no difference (P > 0.05) between grab samples of feces collected in the morning (07:00 h) and evening (19:00 h). In the second period consisting of 22 d, all rumen contents (54–123 kg) were removed at eight different times of the day, sampled and returned at 3-d intervals. Fluctuations in ruminal pool size of small particles (SMALL, kg) were regressed against sampling hour over 24 h. The fitted mean for SMALL was 3.97 ± 0.05 kg with a 0.44 ± 0.02 kg amplitude. On two occasions, chromium-labelled fecal particles were mixed with contents removed from, then returned to the rumen. Marker dilution curves were established up to 72 h after returning the contents to the rumen. Marker concentrations and estimates of SMALL over 48 h were used to compute fluctuations of the fractional rate of passage (kp). The third period served as a control for the INTAKE pattern. INTAKE fitted means, fluctuations and phases were not different (P > 0.05) betweeen periods, and it can be assumed that sequential rumen emptying did not affect the regularity of eating patterns or fluctuations of SMALL. SMALL and INTAKE fluctuations were negatively correlated (r = −0.98, P < 0.01). Largest SMALL occurred almost exactly at the lowest INTAKE, kp and SMALL fluctuations were also negatively correlated (r = −0.68 P < 0.01). Between 19:00 and 05:00 h SMALL increased and kp was less (P < 0.01) than the average rate (0.037 ± 0.004 h−1) during 72 h. When steers were resting or ruminating, small particles remained longer in the rumen than when steers were eating. This should be accounted for in protein degradability predictions. Key words: Particles, rumen, rate of passage, intake, cosinor