Management practices have an effect on nutrient dynamics in grassland ecosystems, and understanding these effects is necessary to discuss the problems of pasture degradation in systems managed extensively. The objective of this study was to evaluate litter C to N ratio in a Marandu-grass pasture managed under three heights of pasture. The heights of 15, 25 and 35 cm were managed by beef steers grazing under continuous stocking. Evaluations of litter decomposition on the soil occurred from Jan 24 to October 16, 2011, at the UNESP, Jaboticabal, SP. The litter bag technique was used in order to evaluate the decomposition of senescent plant material. Incubation times were 4, 8, 16, 32, 64, 128 and 256 days. Nylon bags measuring 15 cm x 30 cm were used with 11.25 g of litter per bag. At the end of each incubation period, the bags were collected and dried at 55ºC for 72 hours, weighed. After this, chemical analyses were done. Nonlinear models were used to fit the curves of  ratio. Data were submitted to variance analysis considering the effects of pasture height and incubation time and the level comparisons of pasture height by incubation time were made using orthogonal polynomial contrasts (p<0.05), using Proc Mixed from SAS. The single exponential model fit data well and was used to develop the C to N ratio curves for each pasture height, being y = 51.001e^-0.004x, y = 45.639e^-0.004x  and y = 45.513e^-0.003x, respectively for 15, 25 and 35 cm. No linear, quadratic or cubic effect was observed (p<0.05) among different treatments for the C to N ratio. Decreasing C to N ratio over time is expected because the more soluble C compounds decompose rapidly, but the N immobilization by the low quality residue and the N bound to the fiber reduce N loss.