Objective The objective is to evaluate the effect of intravenous contrast media on bone mineral density (BMD) assessment by comparing unenhanced and contrast-enhanced computed tomography (CT) examinations performed for other indications. Methods One hundred and fifty-two patients (99 without and 53 with malignant neoplasm) who underwent both unenhanced and two contrast-enhanced (arterial and portal venous phase) abdominal CT examinations in a single session between June 2011 and July 2013 were included. BMD was evaluated on the three examinations as CT-attenuation values in Hounsfield Units (HU) in the first lumbar vertebra (L1). Results CT-attenuation values were significantly higher in both contrast-enhanced phases, compared to the unenhanced phase ( p  < 0.01). In patients without malignancies, mean ± standard deviation (SD) HU-values increased from 128.8 ± 48.6 HU for the unenhanced phase to 142.3 ± 47.2 HU for the arterial phase and 147.0 ± 47.4 HU for the portal phase ( p  < 0.01). In patients with malignancies, HU-values increased from 112.1 ± 38.1 HU to 126.2 ± 38.4 HU and 130.1 ± 37.3 HU ( p  < 0.02), respectively. With different thresholds to define osteoporosis, measurements in the arterial and portal phase resulted in 7-25 % false negatives. Conclusions Our study showed that intravenous contrast injection substantially affects BMD-assessment on CT and taking this into account may improve routine assessment of low BMD in nonquantitative CT. Key Points • Routine CT may gain a role in bone attenuation measurements for osteoporosis • Contrast media injection has substantial influence on CT - derived bone density • Contrast - enhanced CT leads to underestimation of osteoporosis compared to unenhanced CT • Adjusting for contrast injection phase may improve CT screening protocols for osteoporosis .