Olive stones were employed as feedstock for furfural production in two stages: 1) autohydrolysis of hemicellulosic fraction to recover their pentoses, mainly xylose, and 2) subsequent dehydration of pentoses into furfural. Autohydrolysis step was optimized by using different experimental conditions (temperature: 160-200 ºC and time: 30-75 min), giving rise to liquors with different xylose concentrations, since hydrolysis was incomplete in some cases. The combined use of a commercial γ-Al2O3 and CaCl2 led to total hydrolysis of non-hydrolyzed pentosans after autohydrolysis step, and the subsequent dehydration of pentoses into furfural. The maximum values of furfural yield and efficiency were 23 and 96%, respectively, after only 60 minutes at 150 ºC by using liquor obtained by autohydrolysis at 180 ºC and 30 min (L5.1) as source of pentoses. This liquor, L5.1, provided better catalytic results than other liquors which had shown higher xylose concentration after autohydrolysis, probably due to these latter also exhibited a higher concentration of organic acids; thus, the presence of organic acids such as acetic and lactic acid could promote side undesired reactions leading to lower furfural yields. Finally, γ-Al2O3 was more effective for furfural production under these experimental conditions than other solid acid catalysts, such as mesoporous Nb2O5, Nb-doped SBA-15 and Zr-doped HMS silicas, probably due to alumina has a higher density of acid sites.