Abstract: The present study was carried out in Umfakarin natural forest reserve, South Kordofan, Sudan. The main objective was to investigate the possibility of managing Acacia seyal Del. variety seyal (Bernan) for the production of gum talha (gum obtained from Acacia seyal). In this investigation, three stand densities, namely dense, medium and slight, were distinguished based on number of trees per hectare. A total of 482 target trees, covering variable ranges (9-11.5, 13.5-16, 18-20.5 and above 21 cm) of diameter at the base (0.25 m height) were selected for gum tapping experiment. Diameter at breast height of each target tree was measured. Target trees were exposed to tapping, on the first of October, fifteenth of October and first of November, using local tools (sonki and makmak) in addition to untapped trees used as control. The initial collection of gum was completed fifteen days after tapping while the subsequent (7-9 picks) were done in an interval of fifteen days. Gum yielded by each target tree per season was obtained by summing up the gum samples collected from all pickings. Gum production per unit area in each stand density was also determined. Regression tree model and regression analysis under univariate general linear model were used for analysis of the data obtained. Factors affecting gum production were identified. With exception to stand density, findings of the regression tree model revealed that the diameter at breast height, date and tool of tapping are the most important variables influencing gum production. The overall mean of gum yield indicated by the model was 15.35 g/tree/season. Findings of the general linear model showed that the only significant main effects are tree diameter and time of tapping at medium stand density. Significant interaction between the tool (makmak) and the tapping date (first of October) at the same stand density was also detected by the model. Conducting experiments on the production of gum talha in permanent plot trials under different climatic regions of the Sudan is highly recommended.